.env

@@ -1,5 +1,5 @@
 PUBLIC_URL=
-REACT_APP_API_URL=
+REACT_APP_API_URL="/api"
 
 REACT_APP_DEFAULTAUTH=fake
 

.gitignore

@@ -24,3 +24,8 @@ yarn-error.log*
 
 /src/server/label
 /src/server/images
+/src/server/trained_models
+/src/server/model_datasets
+*.log
+src/server/venv/*
+__pycache__/

package.json

@@ -4,12 +4,18 @@
 	"private": true,
 	"dependencies": {
 		"-": "0.0.1",
+		"@mantine/core": "^7.17.1",
+		"@mantine/dropzone": "^7.17.1",
+		"@mantine/hooks": "^7.17.1",
+		"@mantine/notifications": "^7.17.1",
+		"@tabler/icons-react": "^3.30.0",
 		"@testing-library/jest-dom": "^5.17.0",
 		"@testing-library/react": "^13.4.0",
 		"@testing-library/user-event": "^13.5.0",
 		"assert": "^2.1.0",
 		"buffer": "^6.0.3",
 		"child_process": "^1.0.2",
+		"classnames": "^2.5.1",
 		"fs": "0.0.1-security",
 		"numpy": "0.0.1",
 		"opencv-build": "^0.1.9",
@@ -24,7 +30,8 @@
 		"web-vitals": "^2.1.4",
 		"webpack": "^5.96.1",
 		"webpack-cli": "^5.1.4",
-		"webpack-dev-server": "^5.1.0"
+		"webpack-dev-server": "^5.1.0",
+		"zustand": "^5.0.3"
 	},
 	"scripts": {
 		"start": "react-scripts start",
@@ -52,5 +59,10 @@
 	},
 	"opencv4nodejs": {
 		"disableAutoBuild": "1"
+	},
+	"devDependencies": {
+		"postcss-preset-mantine": "^1.17.0",
+		"postcss-simple-vars": "^7.0.1",
+		"tailwindcss": "^3.4.17"
 	}
 }

postcss.config.cjs

@@ -0,0 +1,14 @@
+module.exports = {
+    plugins: {
+      'postcss-preset-mantine': {},
+      'postcss-simple-vars': {
+        variables: {
+          'mantine-breakpoint-xs': '36em',
+          'mantine-breakpoint-sm': '48em',
+          'mantine-breakpoint-md': '62em',
+          'mantine-breakpoint-lg': '75em',
+          'mantine-breakpoint-xl': '88em',
+        },
+      },
+    },
+  };

runs/detect/train/args.yaml

@@ -0,0 +1,106 @@
+task: detect
+mode: train
+model: yolov8n.pt
+data: model_datasets/2025-03-05/data.yaml
+epochs: 50
+time: null
+patience: 10
+batch: 16
+imgsz: 640
+save: true
+save_period: -1
+cache: false
+device: null
+workers: 8
+project: null
+name: train
+exist_ok: false
+pretrained: true
+optimizer: AdamW
+verbose: true
+seed: 0
+deterministic: true
+single_cls: false
+rect: false
+cos_lr: false
+close_mosaic: 10
+resume: false
+amp: true
+fraction: 1.0
+profile: false
+freeze: null
+multi_scale: false
+overlap_mask: true
+mask_ratio: 4
+dropout: 0.0
+val: true
+split: val
+save_json: false
+save_hybrid: false
+conf: null
+iou: 0.7
+max_det: 300
+half: false
+dnn: false
+plots: true
+source: null
+vid_stride: 1
+stream_buffer: false
+visualize: false
+augment: false
+agnostic_nms: false
+classes: null
+retina_masks: false
+embed: null
+show: false
+save_frames: false
+save_txt: false
+save_conf: false
+save_crop: false
+show_labels: true
+show_conf: true
+show_boxes: true
+line_width: null
+format: torchscript
+keras: false
+optimize: false
+int8: false
+dynamic: false
+simplify: true
+opset: null
+workspace: null
+nms: false
+lr0: 0.001
+lrf: 0.01
+momentum: 0.937
+weight_decay: 0.0005
+warmup_epochs: 3.0
+warmup_momentum: 0.8
+warmup_bias_lr: 0.1
+box: 7.5
+cls: 0.5
+dfl: 1.5
+pose: 12.0
+kobj: 1.0
+nbs: 64
+hsv_h: 0.015
+hsv_s: 0.7
+hsv_v: 0.4
+degrees: 0.0
+translate: 0.1
+scale: 0.5
+shear: 0.0
+perspective: 0.0
+flipud: 0.0
+fliplr: 0.5
+bgr: 0.0
+mosaic: 1.0
+mixup: 0.0
+copy_paste: 0.0
+copy_paste_mode: flip
+auto_augment: randaugment
+erasing: 0.4
+crop_fraction: 1.0
+cfg: null
+tracker: botsort.yaml
+save_dir: /home/work/projects/YoLo/runs/detect/train

runs/detect/train2/args.yaml

@@ -0,0 +1,106 @@
+task: detect
+mode: train
+model: yolov8n.pt
+data: model_datasets/2025-03-05/data.yaml
+epochs: 50
+time: null
+patience: 10
+batch: 16
+imgsz: 640
+save: true
+save_period: -1
+cache: false
+device: null
+workers: 8
+project: null
+name: train2
+exist_ok: false
+pretrained: true
+optimizer: AdamW
+verbose: true
+seed: 0
+deterministic: true
+single_cls: false
+rect: false
+cos_lr: false
+close_mosaic: 10
+resume: false
+amp: true
+fraction: 1.0
+profile: false
+freeze: null
+multi_scale: false
+overlap_mask: true
+mask_ratio: 4
+dropout: 0.0
+val: true
+split: val
+save_json: false
+save_hybrid: false
+conf: null
+iou: 0.7
+max_det: 300
+half: false
+dnn: false
+plots: true
+source: null
+vid_stride: 1
+stream_buffer: false
+visualize: false
+augment: false
+agnostic_nms: false
+classes: null
+retina_masks: false
+embed: null
+show: false
+save_frames: false
+save_txt: false
+save_conf: false
+save_crop: false
+show_labels: true
+show_conf: true
+show_boxes: true
+line_width: null
+format: torchscript
+keras: false
+optimize: false
+int8: false
+dynamic: false
+simplify: true
+opset: null
+workspace: null
+nms: false
+lr0: 0.001
+lrf: 0.01
+momentum: 0.937
+weight_decay: 0.0005
+warmup_epochs: 3.0
+warmup_momentum: 0.8
+warmup_bias_lr: 0.1
+box: 7.5
+cls: 0.5
+dfl: 1.5
+pose: 12.0
+kobj: 1.0
+nbs: 64
+hsv_h: 0.015
+hsv_s: 0.7
+hsv_v: 0.4
+degrees: 0.0
+translate: 0.1
+scale: 0.5
+shear: 0.0
+perspective: 0.0
+flipud: 0.0
+fliplr: 0.5
+bgr: 0.0
+mosaic: 1.0
+mixup: 0.0
+copy_paste: 0.0
+copy_paste_mode: flip
+auto_augment: randaugment
+erasing: 0.4
+crop_fraction: 1.0
+cfg: null
+tracker: botsort.yaml
+save_dir: /home/work/projects/YoLo/runs/detect/train2

runs/detect/train3/args.yaml

@@ -0,0 +1,106 @@
+task: detect
+mode: train
+model: yolov8n.pt
+data: model_datasets/2025-03-05/data.yaml
+epochs: 50
+time: null
+patience: 10
+batch: 16
+imgsz: 640
+save: true
+save_period: -1
+cache: false
+device: null
+workers: 8
+project: null
+name: train3
+exist_ok: false
+pretrained: true
+optimizer: AdamW
+verbose: true
+seed: 0
+deterministic: true
+single_cls: false
+rect: false
+cos_lr: false
+close_mosaic: 10
+resume: false
+amp: true
+fraction: 1.0
+profile: false
+freeze: null
+multi_scale: false
+overlap_mask: true
+mask_ratio: 4
+dropout: 0.0
+val: true
+split: val
+save_json: false
+save_hybrid: false
+conf: null
+iou: 0.7
+max_det: 300
+half: false
+dnn: false
+plots: true
+source: null
+vid_stride: 1
+stream_buffer: false
+visualize: false
+augment: false
+agnostic_nms: false
+classes: null
+retina_masks: false
+embed: null
+show: false
+save_frames: false
+save_txt: false
+save_conf: false
+save_crop: false
+show_labels: true
+show_conf: true
+show_boxes: true
+line_width: null
+format: torchscript
+keras: false
+optimize: false
+int8: false
+dynamic: false
+simplify: true
+opset: null
+workspace: null
+nms: false
+lr0: 0.001
+lrf: 0.01
+momentum: 0.937
+weight_decay: 0.0005
+warmup_epochs: 3.0
+warmup_momentum: 0.8
+warmup_bias_lr: 0.1
+box: 7.5
+cls: 0.5
+dfl: 1.5
+pose: 12.0
+kobj: 1.0
+nbs: 64
+hsv_h: 0.015
+hsv_s: 0.7
+hsv_v: 0.4
+degrees: 0.0
+translate: 0.1
+scale: 0.5
+shear: 0.0
+perspective: 0.0
+flipud: 0.0
+fliplr: 0.5
+bgr: 0.0
+mosaic: 1.0
+mixup: 0.0
+copy_paste: 0.0
+copy_paste_mode: flip
+auto_augment: randaugment
+erasing: 0.4
+crop_fraction: 1.0
+cfg: null
+tracker: botsort.yaml
+save_dir: /home/work/projects/YoLo/runs/detect/train3

runs/detect/train3/results.csv

@@ -0,0 +1,3 @@
+epoch,time,train/box_loss,train/cls_loss,train/dfl_loss,metrics/precision(B),metrics/recall(B),metrics/mAP50(B),metrics/mAP50-95(B),val/box_loss,val/cls_loss,val/dfl_loss,lr/pg0,lr/pg1,lr/pg2
+1,60.4197,0,100.785,0,0,0,0,0,0,35.8662,0,0.09901,1e-05,1e-05
+2,119.599,0,62.3699,0,0,0,0,0,0,22.7946,0,0.0970294,2.9406e-05,2.9406e-05

runs/detect/train4/args.yaml

@@ -0,0 +1,106 @@
+task: detect
+mode: train
+model: yolov8n.pt
+data: model_datasets/2025-03-05/data.yaml
+epochs: 2
+time: null
+patience: 10
+batch: 16
+imgsz: 640
+save: true
+save_period: -1
+cache: false
+device: null
+workers: 8
+project: null
+name: train4
+exist_ok: false
+pretrained: true
+optimizer: AdamW
+verbose: true
+seed: 0
+deterministic: true
+single_cls: false
+rect: false
+cos_lr: false
+close_mosaic: 10
+resume: false
+amp: true
+fraction: 1.0
+profile: false
+freeze: null
+multi_scale: false
+overlap_mask: true
+mask_ratio: 4
+dropout: 0.0
+val: true
+split: val
+save_json: false
+save_hybrid: false
+conf: null
+iou: 0.7
+max_det: 300
+half: false
+dnn: false
+plots: true
+source: null
+vid_stride: 1
+stream_buffer: false
+visualize: false
+augment: false
+agnostic_nms: false
+classes: null
+retina_masks: false
+embed: null
+show: false
+save_frames: false
+save_txt: false
+save_conf: false
+save_crop: false
+show_labels: true
+show_conf: true
+show_boxes: true
+line_width: null
+format: torchscript
+keras: false
+optimize: false
+int8: false
+dynamic: false
+simplify: true
+opset: null
+workspace: null
+nms: false
+lr0: 0.001
+lrf: 0.01
+momentum: 0.937
+weight_decay: 0.0005
+warmup_epochs: 3.0
+warmup_momentum: 0.8
+warmup_bias_lr: 0.1
+box: 7.5
+cls: 0.5
+dfl: 1.5
+pose: 12.0
+kobj: 1.0
+nbs: 64
+hsv_h: 0.015
+hsv_s: 0.7
+hsv_v: 0.4
+degrees: 0.0
+translate: 0.1
+scale: 0.5
+shear: 0.0
+perspective: 0.0
+flipud: 0.0
+fliplr: 0.5
+bgr: 0.0
+mosaic: 1.0
+mixup: 0.0
+copy_paste: 0.0
+copy_paste_mode: flip
+auto_augment: randaugment
+erasing: 0.4
+crop_fraction: 1.0
+cfg: null
+tracker: botsort.yaml
+save_dir: /home/work/projects/YoLo/runs/detect/train4

runs/detect/train5/args.yaml

@@ -0,0 +1,106 @@
+task: detect
+mode: train
+model: yolov8n.pt
+data: model_datasets/2025-03-05/data.yaml
+epochs: 2
+time: null
+patience: 10
+batch: 16
+imgsz: 640
+save: true
+save_period: -1
+cache: false
+device: null
+workers: 8
+project: null
+name: train5
+exist_ok: false
+pretrained: true
+optimizer: AdamW
+verbose: true
+seed: 0
+deterministic: true
+single_cls: false
+rect: false
+cos_lr: false
+close_mosaic: 10
+resume: false
+amp: true
+fraction: 1.0
+profile: false
+freeze: null
+multi_scale: false
+overlap_mask: true
+mask_ratio: 4
+dropout: 0.0
+val: true
+split: val
+save_json: false
+save_hybrid: false
+conf: null
+iou: 0.7
+max_det: 300
+half: false
+dnn: false
+plots: true
+source: null
+vid_stride: 1
+stream_buffer: false
+visualize: false
+augment: false
+agnostic_nms: false
+classes: null
+retina_masks: false
+embed: null
+show: false
+save_frames: false
+save_txt: false
+save_conf: false
+save_crop: false
+show_labels: true
+show_conf: true
+show_boxes: true
+line_width: null
+format: torchscript
+keras: false
+optimize: false
+int8: false
+dynamic: false
+simplify: true
+opset: null
+workspace: null
+nms: false
+lr0: 0.001
+lrf: 0.01
+momentum: 0.937
+weight_decay: 0.0005
+warmup_epochs: 3.0
+warmup_momentum: 0.8
+warmup_bias_lr: 0.1
+box: 7.5
+cls: 0.5
+dfl: 1.5
+pose: 12.0
+kobj: 1.0
+nbs: 64
+hsv_h: 0.015
+hsv_s: 0.7
+hsv_v: 0.4
+degrees: 0.0
+translate: 0.1
+scale: 0.5
+shear: 0.0
+perspective: 0.0
+flipud: 0.0
+fliplr: 0.5
+bgr: 0.0
+mosaic: 1.0
+mixup: 0.0
+copy_paste: 0.0
+copy_paste_mode: flip
+auto_augment: randaugment
+erasing: 0.4
+crop_fraction: 1.0
+cfg: null
+tracker: botsort.yaml
+save_dir: /home/work/projects/YoLo/runs/detect/train5

runs/detect/train5/results.csv

@@ -0,0 +1,3 @@
+epoch,time,train/box_loss,train/cls_loss,train/dfl_loss,metrics/precision(B),metrics/recall(B),metrics/mAP50(B),metrics/mAP50-95(B),val/box_loss,val/cls_loss,val/dfl_loss,lr/pg0,lr/pg1,lr/pg2
+1,13.2159,2.49424,4.97029,2.25232,0.0014,0.25,0.00121,0.00012,1.88201,4.79522,2.07429,0.1,0,0
+2,25.7522,2.40923,4.29302,2.0057,0.00145,0.25,0.00124,0.00025,1.98576,4.55288,2.05854,0.0990051,5.05e-06,5.05e-06

src/App.js

@@ -1,13 +1,320 @@
-import logo from "./logo.svg";
+import '@mantine/core/styles.css';
-import "./App.css";
+
-import Main from "./pages/index";
+import './App.css';
+import { Dropzone } from './pages/components/Dropzone';
+
+import { ActionIcon, AppShell, Box, Burger, Button, Group, LoadingOverlay, Progress, ScrollArea, Text, Tooltip } from '@mantine/core';
+import { useDisclosure, useHotkeys } from '@mantine/hooks';
+import { notifications } from '@mantine/notifications';
+import { IconChevronLeft, IconChevronRight, IconImageInPicture, IconRefreshDot, IconTrash } from '@tabler/icons-react';
+import { useEffect, useMemo, useRef, useState } from 'react';
+import { ImageDetect } from './pages/components/ImageDetect';
+import ImageLabel from './pages/components/ImageLabel';
+import { useImagesDetected } from './stores/use-images-detected';
+import { SaveButton } from './pages/components/SaveButton';
+import { generateClientID } from './ultils';
 
 function App() {
+	const [mobileOpened, { toggle: toggleMobile }] = useDisclosure();
+	const [desktopOpened, { toggle: toggleDesktop }] = useDisclosure(true);
+
+	const [clickData, setClickData] = useState(null);
+
+	const { images, clearImagesDetected, appendImageDetect, setImageDetected } = useImagesDetected();
+
+	const [progress, setProgress] = useState({ total: 0, processFiles: 0 });
+
+	const imageLabelRef = useRef();
+	const dropzoneRef = useRef();
+
+	const handleClear = () => {
+		// const result = window.confirm('Are you want to clear ?');
+
+		const ID_NOTI = 'handle-clear';
+		notifications.show({
+			id: ID_NOTI,
+			title: 'Are you want to clear ?',
+			message: (
+				<Box className="flex items-center gap-2 justify-end">
+					<Button
+						onClick={() => {
+							setClickData(null);
+							clearImagesDetected();
+							notifications.hide(ID_NOTI);
+						}}
+						color="red"
+						size="xs">
+						Ok
+					</Button>
+					<Button onClick={() => notifications.hide(ID_NOTI)} size="xs">
+						Cancel
+					</Button>
+				</Box>
+			),
+		});
+	};
+
+	const handleDeleteImage = (data, index) => {
+		const ID_NOTI = 'handle-delete-image';
+
+		const deleteImage = async () => {
+			try {
+				const response = await fetch(`${process.env.REACT_APP_API_URL}/delete/${data.image_name}`, { method: 'DELETE' });
+
+				const result = await response.json();
+
+				setClickData((prevClick) => (prevClick?.image_name === data.image_name ? null : prevClick));
+
+				const newImages = images.filter((image) => image.image_name !== data.image_name);
+				appendImageDetect(newImages);
+
+				notifications.hide(ID_NOTI);
+			} catch (error) {
+				console.log(error);
+			}
+		};
+
+		notifications.show({
+			id: ID_NOTI,
+			title: 'Are you want to delete ?',
+			message: (
+				<Box className="flex items-center gap-2 justify-end">
+					<Button onClick={deleteImage} color="red" size="xs">
+						Ok
+					</Button>
+					<Button onClick={() => notifications.hide(ID_NOTI)} size="xs">
+						Cancel
+					</Button>
+				</Box>
+			),
+		});
+	};
+
+	const handleAddImages = () => {
+		// setClickData(null);
+
+		if (dropzoneRef?.current) {
+			dropzoneRef.current();
+		}
+	};
+
+	const handleClearSelect = () => {
+		setClickData(null);
+	};
+
+	const handleNext = () => {
+		if (!images.length) return;
+
+		if (!clickData) {
+			setClickData(images[0]);
+			return;
+		}
+
+		const findItem = images.find((item) => item?.image_name === clickData?.image_name);
+
+		if (!findItem) return;
+
+		const index = images.indexOf(findItem);
+
+		const nextIndex = index + 1;
+
+		if (index === -1 || nextIndex >= images.length) return;
+
+		console.log(images[nextIndex]);
+		setClickData(images[nextIndex]);
+	};
+
+	const handleUpload = async (files) => {
+		if (!files) return;
+
+		setProgress({
+			total: files.length,
+			processFiles: 0,
+		});
+
+		const worker = new Worker(new URL('./workers/detect-worker.js', import.meta.url));
+
+		worker.onmessage = (event) => {
+			if (event.data.status === 'success') {
+				if (event.data?.result) {
+					setImageDetected(event.data.result);
+					setProgress({
+						total: event.data.total,
+						processFiles: event.data.processFiles,
+					});
+
+					if (event.data.total === event.data.processFiles) {
+						setProgress({
+							total: 0,
+							processFiles: 0,
+						});
+					}
+				}
+			} else {
+				console.error('⚠️ Lỗi:', event.data.error);
+			}
+		};
+
+		worker.postMessage({ action: 'processImages', files });
+	};
+
+	const handlePrev = () => {
+		if (!images.length) return;
+		const findItem = images.find((item) => item?.image_name === clickData?.image_name);
+
+		if (!findItem) return;
+
+		const index = images.indexOf(findItem);
+		const prevIndex = index - 1;
+
+		if (index <= 0) return;
+
+		setClickData(images[prevIndex]);
+	};
+
+	const showNext = useMemo(() => {
+		if (!images.length) return false;
+
+		if (clickData?.image_name === images[images.length - 1]?.image_name) return false;
+
+		return true;
+	}, [images, clickData]);
+
+	const showPrev = useMemo(() => {
+		if (!images.length || !clickData) return false;
+
+		if (clickData?.image_name === images[0]?.image_name) return false;
+
+		return true;
+	}, [images, clickData]);
+
+	useHotkeys([
+		['ArrowRight', handleNext],
+		['ArrowLeft', handlePrev],
+	]);
+
 	return (
-		<div className="App">
+		<div style={{ position: 'relative' }}>
-			<header className="App-header">
+			<AppShell
-				<Main />
+				header={{ height: 60 }}
-			</header>
+				navbar={{
+					width: 300,
+					breakpoint: 'sm',
+					collapsed: { mobile: !mobileOpened, desktop: !desktopOpened },
+				}}
+				padding="md">
+				<AppShell.Header>
+					<Box className="flex items-center justify-between h-full">
+						<Group h="100%" px="md">
+							<Burger opened={mobileOpened} onClick={toggleMobile} hiddenFrom="sm" size="sm" />
+							<Burger opened={desktopOpened} onClick={toggleDesktop} visibleFrom="sm" size="sm" />
+						</Group>
+
+						<Box className="px-4 flex items-center gap-4 w-fit">
+							<Button disabled={!clickData} onClick={handleClearSelect} leftSection={<IconRefreshDot size={14} />} color="orange">
+								Reset select
+							</Button>
+							<SaveButton
+								onSaved={(data) => {
+									handleNext();
+								}}
+								currentData={clickData}
+								imageLabelRef={imageLabelRef}
+							/>
+						</Box>
+					</Box>
+				</AppShell.Header>
+				<AppShell.Navbar p="md">
+					<Box className="flex flex-col justify-between gap-2 w-full h-full">
+						<Box className="flex items-center justify-between gap-4 mb-2">
+							<Button fullWidth disabled={!images.length} onClick={handleClear} leftSection={<IconTrash size={14} />} color="red">
+								Clear
+							</Button>
+							<Button fullWidth disabled={images.length} onClick={handleAddImages} leftSection={<IconImageInPicture size={14} />}>
+								Add images
+							</Button>
+						</Box>
+
+						<ScrollArea h={780} className="pb-5 flex-1">
+							<Box className="flex flex-col gap-3 w-full">
+								{images.length > 0 &&
+									images.map((item, index) => {
+										return (
+											<ImageDetect
+												currentData={clickData}
+												onClick={(data) => {
+													console.log({ data });
+													setClickData(data);
+												}}
+												key={`${item?.name || item?.image_name}_${index}`}
+												file={item}
+												onDelete={(data) => {
+													handleDeleteImage(data, index);
+												}}
+											/>
+										);
+									})}
+							</Box>
+
+							{images.length <= 0 && (
+								<Box className="flex items-center justify-center">
+									<span>No images to process</span>
+								</Box>
+							)}
+						</ScrollArea>
+
+						{progress.total > 0 && (
+							<Box className="flex flex-col items-center justify-end w-full">
+								<Text>
+									{progress.processFiles} / {progress.total}
+								</Text>
+								<Progress className="w-full" value={(progress.processFiles * 100) / progress.total} striped animated />
+							</Box>
+						)}
+					</Box>
+				</AppShell.Navbar>
+				<AppShell.Main style={{ position: 'relative' }}>
+					{clickData && <ImageLabel ref={imageLabelRef} data={clickData} />}
+
+					<Box
+						style={{
+							display: 'flex',
+							alignItems: 'center',
+							justifyContent: 'center',
+							gap: '10px',
+							width: '100%',
+							flexDirection: 'column',
+						}}>
+						<Dropzone
+							openRef={dropzoneRef}
+							hidden={!clickData}
+							onErorrFiles={(errors) => {
+								notifications.show({
+									title: 'Invalid Images',
+									message: `There ${errors.length === 1 ? 'is' : 'are'} ${errors.length} invalid image${
+										errors.length > 1 ? 's' : ''
+									}. Please re-check!`,
+									color: 'red',
+								});
+							}}
+							onFilesChange={handleUpload}
+						/>
+					</Box>
+
+					<Box className="fixed bottom-5 right-5 flex items-center gap-4">
+						<Tooltip label={'Left (<)'}>
+							<ActionIcon disabled={!showPrev} onClick={handlePrev} size={'lg'}>
+								<IconChevronLeft size={20} />
+							</ActionIcon>
+						</Tooltip>
+						<Tooltip label={'Right (>)'}>
+							<ActionIcon disabled={!showNext} onClick={handleNext} size={'lg'}>
+								<IconChevronRight size={20} />
+							</ActionIcon>
+						</Tooltip>
+					</Box>
+				</AppShell.Main>
+			</AppShell>
 		</div>
 	);
 }

src/index.css

@@ -1,13 +1,15 @@
+@tailwind base;
+@tailwind components;
+@tailwind utilities;
+
 body {
 	margin: 0;
-  font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', 'Roboto', 'Oxygen',
+	font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', 'Roboto', 'Oxygen', 'Ubuntu', 'Cantarell', 'Fira Sans', 'Droid Sans', 'Helvetica Neue',
-    'Ubuntu', 'Cantarell', 'Fira Sans', 'Droid Sans', 'Helvetica Neue',
 		sans-serif;
 	-webkit-font-smoothing: antialiased;
 	-moz-osx-font-smoothing: grayscale;
 }
 
 code {
-  font-family: source-code-pro, Menlo, Monaco, Consolas, 'Courier New',
+	font-family: source-code-pro, Menlo, Monaco, Consolas, 'Courier New', monospace;
-    monospace;
 }

src/index.js

@@ -1,14 +1,24 @@
 import React from 'react';
 import ReactDOM from 'react-dom/client';
 import './index.css';
+import '@mantine/dropzone/styles.css';
+import '@mantine/notifications/styles.css';
+import { Notifications } from '@mantine/notifications';
 import App from './App';
 import reportWebVitals from './reportWebVitals';
+import { createTheme, MantineProvider } from '@mantine/core';
+
+const theme = createTheme({
+	/** Put your mantine theme override here */
+});
 
 const root = ReactDOM.createRoot(document.getElementById('root'));
 root.render(
-  <React.StrictMode>
+	<MantineProvider theme={theme} defaultColorScheme="dark">
 		<App />
-  </React.StrictMode>
+
+		<Notifications />
+	</MantineProvider>,
 );
 
 // If you want to start measuring performance in your app, pass a function

src/pages/components/Dropzone.js

@@ -0,0 +1,80 @@
+import { forwardRef } from 'react';
+import { Group, Text } from '@mantine/core';
+import { Dropzone as Dz, IMAGE_MIME_TYPE } from '@mantine/dropzone';
+import { IconPhoto, IconUpload, IconX } from '@tabler/icons-react';
+
+export const Dropzone = forwardRef(({ onErorrFiles, onFilesChange, onReject, hidden, openRef }, ref) => {
+	const handleDrop = async (files) => {
+		const checkImageSize = (file) => {
+			return new Promise((resolve, reject) => {
+				const img = new Image();
+				const objectUrl = URL.createObjectURL(file);
+
+				img.src = objectUrl;
+
+				img.onload = () => {
+					if (img.width >= 500 && img.height >= 500) {
+						resolve(file);
+					} else {
+						reject(`${file.name} is too small (${img.width}x${img.height})`);
+					}
+					URL.revokeObjectURL(objectUrl);
+				};
+
+				img.onerror = () => {
+					reject(`${file.name} is not a valid image.`);
+					URL.revokeObjectURL(objectUrl);
+				};
+			});
+		};
+
+		try {
+			const results = await Promise.allSettled(files.map(checkImageSize));
+
+			const acceptedFiles = results.filter((res) => res.status === 'fulfilled').map((res) => res.value);
+			const errors = results.filter((res) => res.status === 'rejected').map((res) => `❌ ${res.reason}`);
+
+			if (errors.length > 0 && onErorrFiles) {
+				onErorrFiles(errors);
+			}
+
+			if (onFilesChange) {
+				onFilesChange(acceptedFiles, files);
+			}
+		} catch (err) {
+			console.error('Unexpected error:', err);
+		}
+	};
+
+	return (
+		<Dz
+			display={!hidden ? 'none' : 'flex'}
+			openRef={openRef}
+			ref={ref}
+			style={{ width: '100%', height: '400px', display: 'flex', justifyContent: 'center' }}
+			onDrop={handleDrop}
+			onReject={onReject}
+			accept={IMAGE_MIME_TYPE}>
+			<Group justify="center" gap="xl" mih={220} h={'100%'} style={{ pointerEvents: 'none' }}>
+				<Dz.Accept>
+					<IconUpload size={52} color="var(--mantine-color-blue-6)" stroke={1.5} />
+				</Dz.Accept>
+				<Dz.Reject>
+					<IconX size={52} color="var(--mantine-color-red-6)" stroke={1.5} />
+				</Dz.Reject>
+				<Dz.Idle>
+					<IconPhoto size={52} color="var(--mantine-color-dimmed)" stroke={1.5} />
+				</Dz.Idle>
+
+				<div>
+					<Text size="xl" inline>
+						Drag images here or click to select files
+					</Text>
+					<Text size="sm" c="dimmed" inline mt={7}>
+						Attach as many files as you like, each file should not exceed 5mb
+					</Text>
+				</div>
+			</Group>
+		</Dz>
+	);
+});

src/pages/components/ImageDetect.js

@@ -0,0 +1,85 @@
+import { ActionIcon, Avatar, Box, Indicator, LoadingOverlay } from '@mantine/core';
+import { generateImageUrl, randomDelay } from '../../ultils';
+import { useEffect, useState, useCallback, useRef } from 'react';
+import classNames from 'classnames';
+import { useImagesDetected } from '../../stores/use-images-detected';
+import { IconTrash } from '@tabler/icons-react';
+
+export const ImageDetect = ({ file, onClick, currentData, onDelete, onDetected }) => {
+	const [loading, setLoading] = useState(false);
+	const [data, setData] = useState();
+	const [isVisible, setIsVisible] = useState(false);
+	const ref = useRef(null);
+	const [isDetected, setIsDetected] = useState(false);
+
+	useEffect(() => {
+		if (!file?.points) return;
+
+		setData(file);
+	}, [file]);
+
+	// useEffect(() => {
+	// 	if (isVisible || currentData?.image_name !== file?.image_name) return;
+
+	// 	ref.current?.scrollIntoView({
+	// 		behavior: 'smooth',
+	// 		block: 'center',
+	// 		inline: 'nearest',
+	// 	});
+	// }, [currentData, isVisible]);
+
+	useEffect(() => {
+		const observer = new IntersectionObserver(
+			([entry]) => {
+				setIsVisible(entry.isIntersecting);
+			},
+			{ threshold: 1 },
+		);
+
+		if (ref.current) observer.observe(ref.current);
+
+		return () => {
+			if (ref.current) observer.unobserve(ref.current);
+		};
+	}, []);
+
+	return (
+		<Box
+			display={isDetected ? 'none' : 'flex'}
+			ref={ref}
+			onClick={data?.points && onClick ? () => onClick(data) : undefined}
+			className={classNames('flex items-center justify-between gap-3 rounded-md p-3 max-w-full relative', {
+				['hover:bg-gray-900 cursor-pointer']: !loading,
+				['cursor-not-allowed']: loading,
+				['bg-gray-900']: currentData?.image_name === data?.image_name,
+			})}>
+			<Box className="flex items-center gap-3 w-full flex-1 max-w-full">
+				<Indicator color="red" disabled={!!data?.isSave}>
+					<Avatar size={'md'} src={generateImageUrl(data?.image_path)} radius="sm" />
+				</Indicator>
+				<span className="truncate flex-1 w-[140px]">{data?.image_name || ''}</span>
+			</Box>
+			<ActionIcon
+				color="red"
+				onClick={
+					onDelete
+						? (e) => {
+								e.stopPropagation();
+								onDelete(data);
+						  }
+						: undefined
+				}>
+				<IconTrash size={14} />
+			</ActionIcon>
+
+			<LoadingOverlay
+				visible={loading}
+				zIndex={10}
+				overlayProps={{ radius: 'sm', blur: 2 }}
+				loaderProps={{
+					size: 'xs',
+				}}
+			/>
+		</Box>
+	);
+};

src/pages/components/ImageLabel.js

@@ -0,0 +1,71 @@
+import React, { useEffect, useRef, useState, forwardRef, useImperativeHandle, memo } from 'react';
+import { Box, LoadingOverlay } from '@mantine/core';
+import PointsBlur from './PointsBlur';
+import { generateImageUrl } from '../../ultils';
+
+const ImageLabel = forwardRef(({ data }, ref) => {
+	const [imageSrc, setImageSrc] = useState(null);
+	const blurredRegions = useRef([]);
+	const [loaded, setLoaded] = useState(false);
+	const [reloading, setReloading] = useState(false);
+	const [listLabel, setListLabel] = useState([]);
+	const [infoImage, setInfoImage] = useState({});
+	const [imageName, setImageName] = useState('');
+
+	// Expose blurredRegions cho component cha
+	useImperativeHandle(ref, () => ({
+		getBlurredRegions: () => blurredRegions.current,
+		setBlurredRegions: (regions) => {
+			blurredRegions.current = regions;
+		},
+		getSaveData: () => {
+			return {
+				blurredRegions: blurredRegions.current,
+				imageName,
+				infoImage,
+				listLabel,
+			};
+		},
+	}));
+
+	useEffect(() => {
+		if (data?.error) {
+			alert(data.error);
+		} else {
+			if (data?.points) {
+				blurredRegions.current = data?.points?.map((pre) => [
+					{ x: pre.x1, y: pre.y1, label: pre.label },
+					{ x: pre.x2, y: pre.y1, label: pre.label },
+					{ x: pre.x2, y: pre.y2, label: pre.label },
+					{ x: pre.x1, y: pre.y2, label: pre.label },
+				]);
+			}
+			setListLabel(
+				Object.entries(data?.labels).map(([id, name]) => ({
+					id: Number(id),
+					name,
+				})),
+			);
+			setImageSrc(generateImageUrl(data?.image_path));
+			setImageName(data?.image_name);
+			setTimeout(() => setReloading((pre) => !pre), 1000);
+		}
+	}, [data]);
+
+	return (
+		<Box className="w-full h-full flex items-center justify-center relative">
+			<PointsBlur
+				imageSrc={imageSrc}
+				blurredRegions={blurredRegions}
+				loaded={loaded}
+				setLoaded={setLoaded}
+				reloading={reloading}
+				setReloading={setReloading}
+				setInfoImage={setInfoImage}
+				listLabel={listLabel}
+			/>
+		</Box>
+	);
+});
+
+export default memo(ImageLabel);

src/pages/components/PointsBlur.js

@@ -1,4 +1,4 @@
-import React, { useEffect, useRef, useState } from "react";
+import React, { useEffect, useRef, useState } from 'react';
 
 const PointBlurImageComponent = ({ imageSrc, blurredRegions, loaded, setLoaded, reloading, setReloading, setInfoImage, listLabel = [] }) => {
 	const points = useRef([]);
@@ -12,7 +12,7 @@ const PointBlurImageComponent = ({ imageSrc, blurredRegions, loaded, setLoaded,
 
 	useEffect(() => {
 		const canvas = canvasRef.current;
-		const ctx = canvas.getContext("2d");
+		const ctx = canvas.getContext('2d');
 		const image = new Image();
 
 		image.onload = () => {
@@ -23,7 +23,7 @@ const PointBlurImageComponent = ({ imageSrc, blurredRegions, loaded, setLoaded,
 			setInfoImage({ width: image.width, height: image.height });
 		};
 
-		image.crossOrigin = "anonymous";
+		image.crossOrigin = 'anonymous';
 		image.src = imageSrc;
 	}, [imageSrc]);
 
@@ -56,7 +56,7 @@ const PointBlurImageComponent = ({ imageSrc, blurredRegions, loaded, setLoaded,
 				}
 			}
 
-			points.current.push({ x, y, label: listLabel[0]?.name || "cisco" });
+			points.current.push({ x, y, label: listLabel[0]?.name || 'cisco' });
 			drawRedPoint(canvas, x, y);
 
 			if (points.current.length === 4) {
@@ -125,14 +125,14 @@ const PointBlurImageComponent = ({ imageSrc, blurredRegions, loaded, setLoaded,
 			}
 		};
 
-		canvas.addEventListener("mousedown", handleCanvasMouseDown);
+		canvas.addEventListener('mousedown', handleCanvasMouseDown);
-		canvas.addEventListener("mouseup", handleCanvasMouseUp);
+		canvas.addEventListener('mouseup', handleCanvasMouseUp);
-		canvas.addEventListener("mousemove", handleCanvasMouseMove);
+		canvas.addEventListener('mousemove', handleCanvasMouseMove);
 
 		return () => {
-			canvas.removeEventListener("mousedown", handleCanvasMouseDown);
+			canvas.removeEventListener('mousedown', handleCanvasMouseDown);
-			canvas.removeEventListener("mouseup", handleCanvasMouseUp);
+			canvas.removeEventListener('mouseup', handleCanvasMouseUp);
-			canvas.removeEventListener("mousemove", handleCanvasMouseMove);
+			canvas.removeEventListener('mousemove', handleCanvasMouseMove);
 		};
 	}, [draggingPointIndex, movingRegionIndex, offset, isOpen, info]);
 
@@ -160,17 +160,17 @@ const PointBlurImageComponent = ({ imageSrc, blurredRegions, loaded, setLoaded,
 	};
 
 	const drawRedPoint = (canvas, x, y) => {
-		const ctx = canvas.getContext("2d");
+		const ctx = canvas.getContext('2d');
 		ctx.save();
 		ctx.beginPath();
 		ctx.arc(x, y, 5, 0, 2 * Math.PI);
-		ctx.fillStyle = "red";
+		ctx.fillStyle = 'red';
 		ctx.fill();
 		ctx.restore();
 	};
 
 	const drawBorder = (canvas, points) => {
-		const ctx = canvas.getContext("2d");
+		const ctx = canvas.getContext('2d');
 		ctx.save();
 		ctx.beginPath();
 		points.forEach((point, index) => {
@@ -182,7 +182,7 @@ const PointBlurImageComponent = ({ imageSrc, blurredRegions, loaded, setLoaded,
 		});
 		ctx.closePath();
 		ctx.lineWidth = 2;
-		ctx.strokeStyle = "blue";
+		ctx.strokeStyle = 'blue';
 		ctx.stroke();
 		ctx.restore();
 
@@ -190,23 +190,23 @@ const PointBlurImageComponent = ({ imageSrc, blurredRegions, loaded, setLoaded,
 	};
 
 	const drawDeleteButton = (canvas, region, isSelected) => {
-		const ctx = canvas.getContext("2d");
+		const ctx = canvas.getContext('2d');
 		const buttonX = (region[0].x + region[1].x) / 2;
 		const buttonY = Math.min(region[0].y, region[1].y, region[2].y, region[3].y) - 20;
 
 		ctx.save();
-		ctx.fillStyle = isSelected ? "rgb(196, 194, 61)" : "rgb(2, 101, 182)";
+		ctx.fillStyle = isSelected ? 'rgb(196, 194, 61)' : 'rgb(2, 101, 182)';
 		ctx.fillRect(buttonX - 30, buttonY - 10, 60, 20);
-		ctx.fillStyle = "rgb(255, 255, 255)";
+		ctx.fillStyle = 'rgb(255, 255, 255)';
-		ctx.font = "13px Arial";
+		ctx.font = '13px Arial';
-		ctx.textAlign = "center";
+		ctx.textAlign = 'center';
 		ctx.fillText(region[0]?.label, buttonX, buttonY + 5);
 		ctx.restore();
 	};
 
 	const redrawCanvas = (index) => {
 		const canvas = canvasRef.current;
-		const ctx = canvas.getContext("2d");
+		const ctx = canvas.getContext('2d');
 		const image = new Image();
 
 		image.onload = () => {
@@ -223,17 +223,17 @@ const PointBlurImageComponent = ({ imageSrc, blurredRegions, loaded, setLoaded,
 			});
 		};
 
-		image.crossOrigin = "anonymous";
+		image.crossOrigin = 'anonymous';
 		image.src = imageSrc;
 	};
 
 	return (
-		<div style={{ position: "relative" }}>
+		<div style={{ position: 'relative' }}>
 			<canvas ref={canvasRef} />
 			{isOpen && (
-				<div style={{ backgroundColor: "#FFF", position: "absolute", top: info?.value[0].y - 200, left: info?.value[0].x, zIndex: 10 }}>
+				<div style={{ backgroundColor: '#FFF', position: 'absolute', top: info?.value[0].y - 200, left: info?.value[0].x, zIndex: 9999 }}>
 					<div>
-						<span style={{ color: "#000", fontSize: "18px", fontWeight: "bold" }}>Select label</span>
+						<span style={{ color: '#000', fontSize: '18px', fontWeight: 'bold' }}>Select label</span>
 					</div>
 					{listLabel?.map((el, i) => (
 						<div key={i}>
@@ -245,7 +245,7 @@ const PointBlurImageComponent = ({ imageSrc, blurredRegions, loaded, setLoaded,
 									setIsOpen(false);
 									redrawCanvas();
 								}}
-								style={{ backgroundColor: "#ccc", width: "150px", color: "#000" }}>
+								style={{ backgroundColor: '#ccc', width: '150px', color: '#000' }}>
 								{el?.name}
 							</button>
 						</div>
@@ -258,7 +258,7 @@ const PointBlurImageComponent = ({ imageSrc, blurredRegions, loaded, setLoaded,
 								setIsOpen(false);
 								redrawCanvas();
 							}}
-							style={{ backgroundColor: "#ff0000", marginBottom: "10px", width: "150px", color: "#fff" }}>
+							style={{ backgroundColor: '#ff0000', marginBottom: '10px', width: '150px', color: '#fff' }}>
 							Delete
 						</button>
 					</div>

src/pages/components/SaveButton.js

@@ -0,0 +1,114 @@
+import { Button, Tooltip, LoadingOverlay } from '@mantine/core';
+import { notifications } from '@mantine/notifications';
+import { IconImageInPicture } from '@tabler/icons-react';
+import { useState } from 'react';
+import { useImagesDetected } from '../../stores/use-images-detected';
+import { convertToBoundingBox } from '../../ultils';
+import { useHotkeys } from '@mantine/hooks';
+
+export const SaveButton = ({ currentData, imageLabelRef, onSaved }) => {
+	const [loading, setLoading] = useState(false);
+
+	const { appendImageDetect, images } = useImagesDetected();
+
+	const handleSave = async ({ blurredRegions, infoImage, imageName, listLabel }) => {
+		try {
+			if (blurredRegions?.length > 0) {
+				let arrPoints = '';
+				blurredRegions?.forEach((points) => {
+					const img_w = infoImage?.width;
+					const img_h = infoImage?.height;
+
+					// Get bounding box
+					const x_min = Math.min(...points.map((p) => p.x));
+					const y_min = Math.min(...points.map((p) => p.y));
+					const x_max = Math.max(...points.map((p) => p.x));
+					const y_max = Math.max(...points.map((p) => p.y));
+
+					const x_center = (x_min + x_max) / 2 / img_w;
+					const y_center = (y_min + y_max) / 2 / img_h;
+					const width = (x_max - x_min) / img_w;
+					const height = (y_max - y_min) / img_h;
+
+					// Get class ID
+					const class_id = listLabel.find((label) => label?.name === points[0].label)?.id || '0';
+
+					const yolo_label = `${class_id} ${x_center.toFixed(6)} ${y_center.toFixed(6)} ${width.toFixed(6)} ${height.toFixed(6)}`;
+
+					arrPoints += yolo_label + '\n';
+				});
+				const url = `${process.env.REACT_APP_API_URL}/save`;
+				const formData = new FormData();
+				formData.append('list', arrPoints);
+				formData.append('imageName', imageName);
+				setLoading(true);
+				const response = await fetch(url, {
+					method: 'POST',
+					body: formData,
+				});
+
+				const data = await response.json();
+
+				if (data.success) {
+					const newPoints = blurredRegions.map((points) => {
+						return convertToBoundingBox(points);
+					});
+
+					currentData.points = newPoints;
+					currentData['isSave'] = true;
+
+					const newImages = images.map((image) => {
+						if (image.image_name === imageName) {
+							return {
+								...image,
+								points: newPoints,
+								isSave: true,
+							};
+						}
+
+						return { ...image };
+					});
+
+					appendImageDetect(newImages);
+
+					notifications.show({
+						title: 'Save Success',
+						message: `${imageName} save success`,
+						color: 'green',
+					});
+
+					if (onSaved) {
+						onSaved({ ...currentData });
+					}
+				}
+			}
+		} catch (error) {
+			notifications.show({
+				title: 'Save Error',
+				message: error?.message || 'Internal Server Error',
+				color: 'red',
+			});
+		} finally {
+			setLoading(false);
+		}
+	};
+
+	const handleSubmit = () => {
+		const saveData = imageLabelRef.current?.getSaveData();
+
+		if (!saveData) return;
+
+		handleSave(saveData);
+	};
+
+	useHotkeys([['ctrl+S', handleSubmit]]);
+
+	return (
+		<Tooltip label={'Ctrl + S'}>
+			<Button className="relative ư-f" disabled={!currentData} onClick={handleSubmit} leftSection={<IconImageInPicture size={14} />}>
+				Save
+				<LoadingOverlay visible={loading} loaderProps={{ size: 14 }} />
+			</Button>
+		</Tooltip>
+	);
+};

src/pages/index.js

@@ -1,116 +0,0 @@
-import React, { useRef, useState } from "react";
-import PointsBlur from "./components/PointsBlur";
-
-const Main = () => {
-	const [imageSrc, setImageSrc] = useState(null);
-	const blurredRegions = useRef([]);
-	const [loaded, setLoaded] = useState(false);
-	const [reloading, setReloading] = useState(false);
-	const [listLabel, setListLabel] = useState([]);
-	const [infoImage, setInfoImage] = useState({});
-	const [imageName, setImageName] = useState("");
-
-	const handleFileChange = async (event) => {
-		const file = event.target.files[0];
-		if (!file) return;
-
-		const formData = new FormData();
-		formData.append("image", file);
-		blurredRegions.current = [];
-		setLoaded(false);
-		try {
-			const url = "http://localhost:5000/detect_image";
-			const response = await fetch(url, {
-				method: "POST",
-				body: formData,
-			});
-
-			const data = await response.json();
-			if (data.error) {
-				alert(data.error);
-			} else {
-				if (data?.points) {
-					blurredRegions.current = data?.points?.map((pre) => [
-						{ x: pre.x1, y: pre.y1, label: pre.label },
-						{ x: pre.x2, y: pre.y1, label: pre.label },
-						{ x: pre.x2, y: pre.y2, label: pre.label },
-						{ x: pre.x1, y: pre.y2, label: pre.label },
-					]);
-				}
-				setListLabel(
-					Object.entries(data?.labels).map(([id, name]) => ({
-						id: Number(id),
-						name,
-					}))
-				);
-				setImageSrc("http://localhost:5000/" + data?.image_path);
-				setImageName(data?.image_name);
-				setTimeout(() => setReloading((pre) => !pre), 1000);
-				console.log(data);
-			}
-		} catch (error) {
-			console.error("Error detecting QR code:", error);
-		}
-	};
-
-	const handleSave = async () => {
-		if (blurredRegions.current?.length > 0) {
-			let arrPoints = "";
-			blurredRegions.current?.forEach((points) => {
-				const img_w = infoImage?.width;
-				const img_h = infoImage?.height;
-
-				// Get bounding box
-				const x_min = Math.min(...points.map((p) => p.x));
-				const y_min = Math.min(...points.map((p) => p.y));
-				const x_max = Math.max(...points.map((p) => p.x));
-				const y_max = Math.max(...points.map((p) => p.y));
-
-				const x_center = (x_min + x_max) / 2 / img_w;
-				const y_center = (y_min + y_max) / 2 / img_h;
-				const width = (x_max - x_min) / img_w;
-				const height = (y_max - y_min) / img_h;
-
-				// Get class ID
-				const class_id = listLabel.find((label) => label?.name === points[0].label)?.id || "0";
-
-				const yolo_label = `${class_id} ${x_center.toFixed(6)} ${y_center.toFixed(6)} ${width.toFixed(6)} ${height.toFixed(6)}`;
-
-				arrPoints += yolo_label + "\n";
-			});
-			const url = "http://localhost:5000/save";
-			const formData = new FormData();
-			formData.append("list", arrPoints);
-			formData.append("imageName", imageName);
-			await fetch(url, {
-				method: "POST",
-				body: formData,
-			});
-		}
-	};
-
-	return (
-		<div>
-			<div>
-				<input type="file" accept="image/*" onChange={handleFileChange} />
-				<button style={{}} onClick={handleSave}>
-					Save
-				</button>
-			</div>
-			<div>
-				<PointsBlur
-					imageSrc={imageSrc}
-					blurredRegions={blurredRegions}
-					loaded={loaded}
-					setLoaded={setLoaded}
-					reloading={reloading}
-					setReloading={setReloading}
-					setInfoImage={setInfoImage}
-					listLabel={listLabel}
-				/>
-			</div>
-		</div>
-	);
-};
-
-export default Main;

src/server/config.py

@@ -0,0 +1,12 @@
+import os
+
+# Cấu hình thư mục lưu trữ ảnh
+IMAGE_FOLDER = "images"
+LABEL_FOLDER = "label"
+TEMP_FOLDER = "temp"
+TRAINED_MODEL_FOLDER = "trained_models"
+PRETRAINED_MODEL = "train5/weights/best.pt"
+
+# Tạo thư mục nếu chưa có
+for folder in [IMAGE_FOLDER, LABEL_FOLDER, TEMP_FOLDER]:
+    os.makedirs(folder, exist_ok=True)

src/server/label/detect_image_250226_131822.txt

@@ -1,20 +0,0 @@
-1 0.846875 0.679733 0.102500 0.046706
-1 0.399687 0.411593 0.091875 0.032527
-1 0.394375 0.494579 0.096250 0.036697
-1 0.380000 0.703086 0.107500 0.041701
-1 0.384062 0.592160 0.104375 0.041701
-0 0.751563 0.582152 0.038125 0.031693
-0 0.302187 0.718098 0.043125 0.031693
-0 0.324688 0.505421 0.038125 0.028357
-0 0.725938 0.486239 0.036875 0.028357
-0 0.334375 0.416597 0.035000 0.025855
-0 0.774687 0.688907 0.039375 0.036697
-1 0.399375 0.333194 0.090000 0.032527
-0 0.310000 0.603003 0.038750 0.031693
-0 0.697187 0.325271 0.033125 0.020017
-0 0.703125 0.401168 0.032500 0.025021
-0 0.336875 0.337364 0.035000 0.022519
-1 0.792500 0.481234 0.092500 0.033361
-1 0.820937 0.572560 0.100625 0.040867
-1 0.767188 0.397832 0.075000 0.023353
-1 0.757813 0.321518 0.073750 0.027523

src/server/label/detect_image_250226_142652.txt

@@ -1,20 +0,0 @@
-1 0.846875 0.679733 0.102500 0.046706
-1 0.399687 0.411593 0.091875 0.032527
-1 0.394375 0.494579 0.096250 0.036697
-1 0.380000 0.703086 0.107500 0.041701
-1 0.384062 0.592160 0.104375 0.041701
-0 0.751563 0.582152 0.038125 0.031693
-0 0.302187 0.718098 0.043125 0.031693
-0 0.324688 0.505421 0.038125 0.028357
-0 0.725938 0.486239 0.036875 0.028357
-0 0.334375 0.416597 0.035000 0.025855
-0 0.773438 0.689741 0.039375 0.036697
-1 0.401875 0.336530 0.090000 0.032527
-0 0.310000 0.603003 0.038750 0.031693
-0 0.697187 0.325271 0.033125 0.020017
-0 0.703125 0.401168 0.032500 0.025021
-1 0.792500 0.481234 0.092500 0.033361
-1 0.820937 0.572560 0.100625 0.040867
-1 0.759687 0.325271 0.080000 0.028357
-1 0.766250 0.399917 0.076875 0.032527
-1 0.337187 0.341952 0.031250 0.026689

src/server/label/detect_image_250226_160719.txt

@@ -1,20 +0,0 @@
-1 0.726563 0.259766 0.156250 0.050781
-0 0.503418 0.502604 0.045898 0.093750
-0 0.444336 0.773438 0.050781 0.098958
-1 0.324219 0.819661 0.164063 0.061198
-1 0.787109 0.660156 0.162109 0.054688
-1 0.387207 0.535156 0.166992 0.049479
-1 0.795410 0.522135 0.159180 0.052083
-1 0.740723 0.800781 0.151367 0.054688
-0 0.455078 0.638672 0.046875 0.095052
-1 0.769531 0.389974 0.154297 0.050781
-1 0.333984 0.677083 0.167969 0.054688
-1 0.321777 0.264323 0.161133 0.046875
-0 0.853027 0.755859 0.047852 0.092448
-1 0.354492 0.398438 0.166016 0.046875
-0 0.473145 0.367839 0.045898 0.092448
-0 0.905273 0.487630 0.042969 0.095052
-0 0.897949 0.620443 0.043945 0.097656
-0 0.837402 0.231120 0.043945 0.095052
-0 0.439941 0.236979 0.045898 0.093750
-1 0.908691 0.358073 0.051758 0.093750

src/server/main.py

@@ -1,93 +1,22 @@
-from flask import Flask, request, jsonify, send_from_directory
+from flask import Flask
 from flask_cors import CORS
-import cv2
+from routes.detect import detect_bp
-import numpy as np
+from routes.save import save_bp
-from pyzbar.pyzbar import decode
+from routes.show import show_bp
-from ultralytics import YOLO
+from routes.delete import delete_bp
-import datetime
+from routes.upload import upload_bp
-from PIL import Image
-import os
-
-# Load model đã huấn luyện
-model = YOLO("train5/weights/best.pt")
 
 app = Flask(__name__)
 CORS(app)
 
-# Ensure the directory exists
-os.makedirs("label", exist_ok=True)
-os.makedirs("images", exist_ok=True)
 
-@app.route('/detect_image', methods=['POST'])
+app.config["MAX_CONTENT_LENGTH"] = 100 * 1024 * 1024 
-def detect_image():
-    file = request.files.get('image')
-    if not file:
-        return jsonify({"error": "No file uploaded"}), 400
 
-    # Open the image using PIL
+app.register_blueprint(detect_bp)
-    try:
+app.register_blueprint(save_bp)
-        image = Image.open(file)
+app.register_blueprint(show_bp)
-    except Exception as e:
+app.register_blueprint(delete_bp)
-        return jsonify({"error": "Invalid image file", "details": str(e)}), 400
+app.register_blueprint(upload_bp)
-
-    # Check image size
-    min_width, min_height = 500, 500
-    if image.width < min_width or image.height < min_height:
-        return jsonify({"error": f"Image is too small. Minimum size is {min_width}x{min_height} pixels."}), 400
-
-    # Reset file pointer before saving
-    file.seek(0)
-
-    basename = "detect_image"
-    suffix = datetime.datetime.now().strftime("%y%m%d_%H%M%S") 
-    image_name = "_".join([basename, suffix]) # e.g. 'detect_image_120508_171442'
-    image_path = os.path.join("images", image_name) + ".png"
-    file.save(image_path)
-    # Chạy mô hình để dự đoán
-    results = model(image_path, conf=0.6)  # Hạ conf xuống để giữ nhiều phát hiện hơn
-    points = []
-    # Xử lý kết quả dự đoán
-    for r in results:
-        for box in r.boxes:
-            x1, y1, x2, y2 = map(int, box.xyxy[0])  # Lấy tọa độ bbox
-            # print(x1, y1, x2, y2)
-            # print(model.names)
-            conf = float(box.conf[0])  # Độ tin cậy
-            cls = int(box.cls[0])  # Nhãn class ID
-            # label = f"{model.names[cls]} {conf:.2f}"  # Tạo nhãn hiển thị
-
-            points.append({'label': model.names[cls],'x1':x1, "y1":y1, "x2":x2, "y2":y2})
-    result = {
-        "image_name": image_name,
-        "image_path": image_path,
-        "labels": model.names,
-        "points": points
-    }
-    return jsonify(result)
-
-# Serve images from the "images" directory
-@app.route('/images/<filename>')
-def get_image(filename):
-    return send_from_directory("images", filename)
-
-@app.route('/save', methods=['POST'])
-def save_data():
-    # Get form data
-    arr_points = request.form.get("list")
-    image_name = request.form.get("imageName")
-
-    if not arr_points or not image_name:
-        return jsonify({"error": "Missing required fields"}), 400
-
-    # Define the file path
-    file_path = os.path.join("label", f"{image_name}.txt")
-    arr_points=arr_points.replace("\r\n", "\n")
-    # Save the data to a text file
-    with open(file_path, "w") as file:
-        file.write(arr_points)  # Write the points directly to the file
-
-
-    return jsonify({"success": True, "list": arr_points, "imageName": image_name})
 
 if __name__ == '__main__':
     app.run(debug=True)

src/server/note.md

@@ -0,0 +1,9 @@
+```bash
+
+   nohup gunicorn -w 1 -b 127.0.0.1:5000 main:app > output.log 2>&1 &
+   ps aux | grep gunicorn
+
+   flask --app main.py --debug run
+
+   source venv/bin/activate.fish
+```

src/server/routes/delete.py

@@ -0,0 +1,24 @@
+import os
+import datetime
+from flask import Blueprint, request, jsonify
+from config import LABEL_FOLDER, IMAGE_FOLDER
+delete_bp = Blueprint('delete', __name__)
+
+@delete_bp.route('/delete/<image_name>', methods=['DELETE'])
+def delete_image(image_name):
+
+    today_str = datetime.datetime.now().strftime("%Y-%m-%d")
+
+    image_path = os.path.join(IMAGE_FOLDER,today_str, image_name)
+    
+    label_path = os.path.join(LABEL_FOLDER,today_str, f"{image_name}.txt")
+
+    if os.path.exists(image_path):
+        os.remove(image_path)
+    else:
+        return jsonify({"error": "Image not found"}), 404
+
+    if os.path.exists(label_path):
+        os.remove(label_path)
+
+    return jsonify({"message": "Image and label deleted successfully", "status": "true"})

src/server/routes/detect.py

@@ -0,0 +1,144 @@
+import os
+import datetime
+from flask import Blueprint, request, jsonify, Response, stream_with_context
+from PIL import Image
+from ultralytics import YOLO
+from config import IMAGE_FOLDER, PRETRAINED_MODEL, TRAINED_MODEL_FOLDER, TEMP_FOLDER
+from utils import ensure_correct_permissions, get_latest_model_including_today
+import random
+import cv2
+import json
+import time
+
+# Load mô hình YOLO
+model = YOLO(get_latest_model_including_today(trained_model_folder=TRAINED_MODEL_FOLDER))
+
+detect_bp = Blueprint('detect', __name__)
+
+
+connected_clients = {}  # Lưu client_id và kết nối
+
+@detect_bp.route('/detect_images', methods=['POST'])
+def detect_images():
+    files = request.files.getlist('images')  # Nhận nhiều ảnh từ request
+    if not files or len(files) == 0:
+        return jsonify({"error": "No files uploaded"}), 400
+
+    results_list = []  # Danh sách kết quả cho từng ảnh
+
+
+    for file in files:
+        try:
+            # Kiểm tra file hợp lệ không
+            image = Image.open(file)
+
+            # Check kích thước ảnh
+            min_width, min_height = 500, 500
+            if image.width < min_width or image.height < min_height:
+                results_list.append({"filename": file.filename, "error": f"Image too small ({image.width}x{image.height})"})
+                continue  # Bỏ qua ảnh lỗi
+
+            # Reset file pointer trước khi lưu
+            file.seek(0)
+
+            # Tạo tên file duy nhất
+            suffix = datetime.datetime.now().strftime("%y%m%d_%H%M%S")
+            image_name = f"detect_{suffix}_{file.filename}"
+            image_path = os.path.join("images", image_name)
+            file.save(image_path)
+
+            # Chạy mô hình YOLO để phát hiện vật thể
+            results = model(image_path, conf=0.6)
+
+            points = []
+            for r in results:
+                for box in r.boxes:
+                    x1, y1, x2, y2 = map(int, box.xyxy[0])  # Lấy tọa độ bbox
+                    conf = float(box.conf[0])  # Độ tin cậy
+                    cls = int(box.cls[0])  # Nhãn class ID
+                    points.append({'label': model.names[cls], 'x1': x1, "y1": y1, "x2": x2, "y2": y2})
+
+            results_list.append({
+                "filename": file.filename,
+                "image_name": image_name,
+                "image_path": image_path,
+                "labels": model.names,
+                "points": points
+            })
+
+        except Exception as e:
+            results_list.append({"filename": file.filename, "error": str(e)})
+
+    return jsonify({"results": results_list})
+
+
+@detect_bp.route('/detect_image', methods=['POST'])
+def detect_image():
+    file = request.files.get('image')
+    if not file:
+        return jsonify({"error": "No file uploaded"}), 400
+
+    # Open the image using PIL
+    try:
+        image = Image.open(file)
+    except Exception as e:
+        return jsonify({"error": "Invalid image file", "details": str(e)}), 400
+
+    # Check image size
+    min_width, min_height = 500, 500
+    if image.width < min_width or image.height < min_height:
+        return jsonify({"error": f"Image is too small. Minimum size is {min_width}x{min_height} pixels."}), 400
+
+    # Reset file pointer before saving
+    file.seek(0)
+
+    # 📌 Get time today
+    today_str = datetime.datetime.now().strftime("%Y-%m-%d")
+
+    # 📌 Create folder images/today if not exits
+    save_dir = os.path.join(IMAGE_FOLDER, today_str)
+    os.makedirs(save_dir, exist_ok=True)
+    ensure_correct_permissions(save_dir)
+    # 📌 Create radom image name
+    basename = "detect_image"
+    suffix = datetime.datetime.now().strftime("%H%M%S")  # Chỉ lấy giờ phút giây
+    random_number = random.randint(100000, 999999)
+    image_name = f"{basename}_{suffix}_{random_number}.png"
+    image_path = os.path.join(save_dir, image_name)
+
+    # 📌 Save image to folder today time
+    file.save(image_path)
+
+    # Run with model
+    results = model(image_path, conf=0.6)  
+    points = []
+
+    # Result predict
+    for r in results:
+        for box in r.boxes:
+            x1, y1, x2, y2 = map(int, box.xyxy[0])  
+            conf = float(box.conf[0])  
+            cls = int(box.cls[0])  
+
+            points.append({'label': model.names[cls], 'x1': x1, "y1": y1, "x2": x2, "y2": y2})
+
+    result = {
+        "image_name": image_name,
+        "image_path": image_path,
+        "labels": model.names,
+        "points": points
+    }
+    return jsonify(result)
+
+
+@detect_bp.route('/reload_model', methods=['POST'])
+def reload_model():
+    global model
+    new_model_path = get_latest_model_including_today(trained_model_folder=TRAINED_MODEL_FOLDER)
+    print(new_model_path)
+    model = YOLO(new_model_path)  
+    return jsonify({"message": f"Model reloaded from {new_model_path}", "status": "true"}), 200
+
+
+
+

src/server/routes/save.py

@@ -0,0 +1,29 @@
+import os
+from flask import Blueprint, request, jsonify
+from config import LABEL_FOLDER
+import datetime
+from utils import ensure_correct_permissions
+save_bp = Blueprint('save', __name__)
+
+@save_bp.route('/save', methods=['POST'])
+def save_data():
+    arr_points = request.form.get("list")
+    image_name = request.form.get("imageName")
+
+    if not arr_points or not image_name:
+        return jsonify({"error": "Missing required fields"}), 400
+
+    today_str = datetime.datetime.now().strftime("%Y-%m-%d")
+
+    today_folder = os.path.join(LABEL_FOLDER, today_str)
+    os.makedirs(today_folder, exist_ok=True)
+    ensure_correct_permissions(today_folder)  
+
+    file_name = os.path.splitext(image_name)[0] 
+    file_path = os.path.join(today_folder, f"{file_name}.txt")
+    arr_points = arr_points.replace("\r\n", "\n")
+
+    with open(file_path, "w") as file:
+        file.write(arr_points)
+
+    return jsonify({"success": True, "list": arr_points, "imageName": image_name})

src/server/routes/show.py

@@ -0,0 +1,17 @@
+import os
+import datetime
+from flask import Blueprint, request, jsonify, send_from_directory
+from PIL import Image
+from ultralytics import YOLO
+from config import IMAGE_FOLDER
+
+show_bp = Blueprint('show', __name__)
+
+@show_bp.route('/images/<date>/<filename>')
+def get_image(date, filename):
+    image_path = os.path.join("images", date)
+
+    if not os.path.exists(image_path):
+        return jsonify({"error": "Image not found"}), 404
+
+    return send_from_directory(image_path, filename)

src/server/routes/upload.py

@@ -0,0 +1,31 @@
+import os
+from flask import Blueprint, request, jsonify
+from config import TEMP_FOLDER
+import datetime
+from utils import ensure_correct_permissions
+upload_bp = Blueprint('upload', __name__)
+
+@upload_bp.route('/upload_images', methods=['POST'])
+def upload_images():
+    files = request.files.getlist("images")
+    if not files:
+        return jsonify({"error": "No files received"}), 400
+
+    saved_files = []
+    skipped_files = []
+
+    for file in files:
+        file_path = os.path.join(TEMP_FOLDER, file.filename)
+
+        if os.path.exists(file_path):  
+            skipped_files.append(file.filename)
+            continue
+
+        file.save(file_path)
+        saved_files.append(file.filename)
+
+    return jsonify({
+        "message": "Upload completed!",
+        "saved": saved_files,
+        "skipped": skipped_files
+    }), 200

src/server/train.py

@@ -0,0 +1,235 @@
+import os
+import shutil
+import datetime
+import random
+from config import IMAGE_FOLDER, LABEL_FOLDER, TRAINED_MODEL_FOLDER, PRETRAINED_MODEL
+from ultralytics import YOLO
+import yaml
+import requests
+import glob
+
+# 🗂️ Configure paths
+DATA_SPLIT_FOLDER = "model_datasets"
+IMAGE_EXTENSION = ".png"  
+LOG_FILE = "training_logs.log"
+
+class_names = ["cisco", "barcode"]
+# 📅 Get today's and yesterday's date
+today = datetime.date.today()
+yesterday = today - datetime.timedelta(days=1)
+
+today_str = today.strftime("%Y-%m-%d")
+yesterday_str = yesterday.strftime("%Y-%m-%d")
+
+image_folder_today = os.path.join(IMAGE_FOLDER, today_str)
+image_folder_yesterday = os.path.join(IMAGE_FOLDER, yesterday_str)
+
+label_folder_today = os.path.join(LABEL_FOLDER, today_str)
+label_folder_yesterday = os.path.join(LABEL_FOLDER, yesterday_str)
+
+# 🔍 Validate folder existence
+if os.path.exists(image_folder_today) and os.path.exists(label_folder_today):
+    image_folder = image_folder_today
+    label_folder = label_folder_today
+    date_str = today_str
+elif os.path.exists(image_folder_yesterday) and os.path.exists(label_folder_yesterday):
+    image_folder = image_folder_yesterday
+    label_folder = label_folder_yesterday
+    date_str = yesterday_str
+else:
+    print(image_folder)
+    print(label_folder)
+    raise Exception("⚠️ No valid image & label folder found in the last two days!")
+
+# 🏗️ Create dataset folders
+dataset_folder = os.path.join(DATA_SPLIT_FOLDER, date_str)
+model_folder_name = os.path.join(TRAINED_MODEL_FOLDER)
+nolable_img_folder = os.path.join(dataset_folder, "nolable")
+train_img_folder = os.path.join(dataset_folder, "train", "images")
+train_lbl_folder = os.path.join(dataset_folder, "train", "labels")
+val_img_folder = os.path.join(dataset_folder, "val", "images")
+val_lbl_folder = os.path.join(dataset_folder, "val", "labels")
+
+for folder in [train_img_folder, train_lbl_folder, val_img_folder, val_lbl_folder, nolable_img_folder]:
+    os.makedirs(folder, exist_ok=True)
+
+# 📂 Get image list
+image_files = [f for f in os.listdir(image_folder) if f.endswith(IMAGE_EXTENSION)]
+
+# 🌀 Shuffle images
+random.shuffle(image_files)
+
+# 📊 Split ratio
+train_ratio = 0.8
+split_idx = int(len(image_files) * train_ratio)
+
+train_files = image_files[:split_idx]
+val_files = image_files[split_idx:]
+
+
+
+def log_message(message: str):
+    """Ghi log vào file"""
+    timestamp = datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+    log_entry = f"[{timestamp}] {message}\n"
+    with open(LOG_FILE, "a") as log:
+        log.write(log_entry)
+
+def get_latest_model(trained_model_folder: str, default_model: str = "train5/weights/best.pt"):
+    """Tìm model gần nhất (trước ngày hôm nay), nếu không có thì dùng model mặc định."""
+    if not os.path.exists(trained_model_folder):
+        log_message(f"⚠️ Folder {trained_model_folder} does not exist. Using default model: {default_model}")
+        return default_model
+
+    today_str = datetime.datetime.now().strftime("%Y-%m-%d")
+
+    # Lấy danh sách các thư mục theo ngày, loại bỏ thư mục của ngày hôm nay
+    subfolders = [
+        f for f in os.listdir(trained_model_folder) 
+        if os.path.isdir(os.path.join(trained_model_folder, f)) and f < today_str
+    ]
+    subfolders = sorted(subfolders, reverse=True)  # Sắp xếp giảm dần (mới nhất trước)
+
+    for folder in subfolders:
+        model_path = os.path.join(trained_model_folder, folder, "weights", "best.pt")
+        if os.path.exists(model_path):
+            log_message(f"✅ Found latest model: {model_path}")
+            return model_path
+
+    log_message(f"⚠️ No valid models found in {trained_model_folder}. Using default model: {default_model}")
+    return default_model
+
+def create_dataset_yaml(dataset_folder: str, classes: list):
+    dataset_path = os.path.abspath(dataset_folder)  
+    data_yaml_path = os.path.join(dataset_folder, "data.yaml")
+
+    data = {
+        "train": os.path.join(dataset_path, "train", "images"),
+        "val": os.path.join(dataset_path, "val", "images"),
+        "test": os.path.join(dataset_path, "test", "images"),  # Nếu có test set
+        "nc": len(classes),
+        "names": list(classes)
+    }
+
+    with open(data_yaml_path, "w") as f:
+        yaml.dump(data, f, default_flow_style=True)
+
+    return data_yaml_path
+
+def call_reload_model_api(base_url="http://localhost:5000"):
+    """Gọi API để reload model"""
+    url = f"{base_url}/reload_model"
+    try:
+        response = requests.post(url, timeout=10)
+        response_data = response.json()
+        log_message(f"✅ API Response: {response_data} \n")
+        return response_data
+    except requests.RequestException as e:
+        log_message(f"❌ Error calling reload model API: {e} \n")
+        return {"error": str(e)}
+
+def clear_images_source():
+    paths = [image_folder, label_folder]
+    # paths = [image_folder, label_folder, train_img_folder, train_lbl_folder, val_img_folder, val_lbl_folder]
+
+    for path in paths:
+        for file in glob.glob(os.path.join(path, "*")):
+            if os.path.isfile(file):  
+                os.remove(file)
+
+    
+    log_message(f"Delete source image success \n")
+    log_message("END " + ("=" * 20) + "\n")
+
+
+# 🚀 Function to move files and log process
+def move_files(file_list, dest_img_folder, dest_lbl_folder):
+    copied_images = 0
+    copied_labels = 0
+    
+    with open(LOG_FILE, "a") as log:
+        for img_file in file_list:
+            img_path = os.path.join(image_folder, img_file)
+            label_file = os.path.splitext(img_file)[0] + ".txt"
+            label_path = os.path.join(label_folder, label_file)
+            
+            if os.path.exists(img_path) and os.path.exists(label_path):
+                shutil.copy(img_path, os.path.join(dest_img_folder, img_file))
+                shutil.copy(label_path, os.path.join(dest_lbl_folder, label_file))
+                copied_images += 1
+                copied_labels += 1
+            else:
+                log.write(f"[{datetime.datetime.now()}] ⚠️ Missing label for image: {img_file}\n")
+                shutil.copy(img_path, os.path.join(nolabel_img_folder, img_file))
+                
+                
+    
+    return copied_images, copied_labels
+
+
+def train_yolo_model(pretrained_model: str, dataset_folder: str,project_name: str, name: str, epochs: int = 50, batch_size: int = 16, img_size: int = 640, lr: float = 0.001):
+    dataset_yaml = os.path.join(dataset_folder, "data.yaml")
+
+    if not os.path.exists(dataset_yaml):
+        raise FileNotFoundError(f"⚠️ Not found file {dataset_yaml}. Plases check datasets!")
+
+    if not os.path.exists(pretrained_model):
+        log_message(f"⚠️ Model not found {pretrained_model}. Start train with 'yolov8n.pt'.")
+        pretrained_model = "yolov8n.pt"
+
+    # 🚀 Tạo model YOLOv8 và load model đã train trước đó
+    model = YOLO(pretrained_model)
+
+    # 🔥 Bắt đầu training
+    model.train(
+        data=dataset_yaml,
+        epochs=epochs,
+        batch=batch_size,
+        imgsz=img_size,
+        optimizer="AdamW",
+        lr0=lr,
+        weight_decay=0.0005,
+        patience=10, 
+        verbose=True,
+        project=project_name,  
+        name=name  
+    )
+
+    with open(LOG_FILE, "a") as log:
+        log.write(f"\n[{datetime.datetime.now()}] ✅ Train completed\n")
+
+    call_reload_model_api()
+    clear_images_source()
+ 
+
+
+
+# 🚀 Copy files
+train_copied_imgs, train_copied_lbls = move_files(train_files, train_img_folder, train_lbl_folder)
+val_copied_imgs, val_copied_lbls = move_files(val_files, val_img_folder, val_lbl_folder)
+
+# Create yml file
+create_dataset_yaml(dataset_folder, class_names)
+
+
+# 🏁 Log summary
+with open(LOG_FILE, "a") as log:
+    log.write(f"\n[{datetime.datetime.now()}] ✅ Dataset split completed\n")
+    log.write(f"Source folder: {image_folder}\n")
+    log.write(f"Dataset saved in: {dataset_folder}\n")
+    log.write(f"Planned Train: {len(train_files)} images, Val: {len(val_files)} images\n")
+    log.write(f"Actual Train: {train_copied_imgs} images, {train_copied_lbls} labels\n")
+    log.write(f"Actual Val: {val_copied_imgs} images, {val_copied_lbls} labels\n")
+
+
+if(train_copied_imgs <=0 or train_copied_lbls <= 0 or val_copied_imgs <=0 or val_copied_lbls <=0):
+    with open(LOG_FILE, "a") as log:
+        log.write(f"\n[{datetime.datetime.now()}] ❌ Data not qualified\n")
+        log.write("=" * 50 + "\n")
+else:
+    train_yolo_model(pretrained_model = get_latest_model(
+        trained_model_folder=TRAINED_MODEL_FOLDER, 
+        default_model=PRETRAINED_MODEL), 
+        dataset_folder = dataset_folder, epochs = 2, name=today_str, project_name=model_folder_name
+        )
+

src/server/train.sh

@@ -0,0 +1,24 @@
+#!/bin/bash
+
+VENV_PATH="/home/work/projects/YoLo/src/server/venv"
+PROJECT_DIR="/home/work/projects/YoLo/src/server"
+TRAIN_SCRIPT="train.py"
+LOCK_FILE="/tmp/train_model.lock"
+LOG_FILE="$PROJECT_DIR/train.log"
+
+{
+    echo "🚀 $(date) - Start running train.py"
+
+    # Chuyển vào thư mục làm việc
+    cd "$PROJECT_DIR" || { echo "❌ Failed to cd into $PROJECT_DIR"; exit 1; }
+
+    echo "📌 Current directory: $(pwd)"
+    
+    # Kích hoạt môi trường ảo
+    source "$VENV_PATH/bin/activate"
+
+    # Chạy script training
+    python3 "$TRAIN_SCRIPT"
+
+    echo "✅ $(date) - Training completed!"
+} 2>&1 | flock -n "$LOCK_FILE" tee -a "$LOG_FILE"

src/server/train/classes.txt

@@ -1,2 +0,0 @@
-0
-1

src/server/train/detect_image_250226_112327.txt

@@ -1,2 +0,0 @@
-0 0.304375 0.714345 0.043750 0.032527
-1 0.380937 0.706422 0.095625 0.031693