{"id":23624,"date":"2024-12-11T11:11:06","date_gmt":"2024-12-11T03:11:06","guid":{"rendered":"https:\/\/blog.tibame.com\/?p=23624"},"modified":"2025-02-24T19:14:15","modified_gmt":"2025-02-24T11:14:15","slug":"yolo-%e6%8a%80%e8%a1%93%e5%85%a8%e9%9d%a2%e5%8d%87%e7%b4%9a-%e5%be%9e-yolov7-%e5%88%b0-yolov10-%e7%9a%84%e7%aa%81%e7%a0%b4%e8%88%87%e6%9c%80%e6%96%b0%e7%99%bc%e5%b1%95","status":"publish","type":"post","link":"https:\/\/blog.tibame.com\/?p=23624","title":{"rendered":"YOLO \u6280\u8853\u5168\u9762\u5347\u7d1a ! \u5f9e YOLOv7 \u5230 YOLOv10 \u7684\u7a81\u7834\u8207\u6700\u65b0\u767c\u5c55"},"content":{"rendered":"\n

\u64b0\u6587\u8005\uff0f\u6797\u54f2\u8070 Alex \u8001\u5e2b <\/p>\n\n\n\n

\u5728\u96fb\u8166\u8996\u89ba\u4e2d\uff0c\u5ee3\u610f\u7684\u5f71\u50cf\u8fa8\u8b58\u5176\u5be6\u9084\u5305\u542b\u8457\u8a31\u591a\u4e0d\u540c\u7684\u4efb\u52d9\uff0c\u5176\u4e2d\u7269\u9ad4\u5075\u6e2c \uff08Object Detection\uff09\u662f\u6700\u5ee3\u6cdb\u7684\u4e00\u7a2e\u61c9\u7528\u4e4b\u4e00\uff0c\u81ea\u5f9e 2015 \u5e74 YOLOv1 \u6a6b\u7a7a\u51fa\u4e16\uff0c\u5728\u6df1\u5ea6\u5b78\u7fd2\u7684\u4e16\u754c\u4e2d\uff0c\u8a31\u591a\u4eba\u767c\u73fe\u6a21\u578b\u8fa8\u8b58\u7387\u4e0d\u4e00\u5b9a\u8981\u8ffd\u6c42\u6975\u70ba\u6e96\u78ba\uff0c\u4f46\u901f\u5ea6\u4e00\u5b9a\u8981\u5feb\uff0c\u4f8b\u5982\u5728\u81ea\u99d5\u8eca\u7684\u61c9\u7528\u4e2d\uff0c\u53d7\u9650\u65bc\u8eca\u8f09\u786c\u9ad4\u6548\u80fd\u4e0a\u7684\u9650\u5236\uff0c\u63a8\u8ad6\u901f\u5ea6\u662f\u6700\u91cd\u8981\u7684\uff0c\u800c\u53ea\u8981\u63d0\u5347\u786c\u9ad4\u4e00\u500b\u7b49\u7d1a\uff0c\u610f\u5473\u8457\u88fd\u9020\u5546\u7684\u7372\u5229\u5c31\u6703\u9650\u4f4e\uff0c\u56e0\u6b64\uff0cYOLO \u4e00\u76f4\u662f\u6709\u5373\u6642\u904b\u7b97\u9700\u6c42\u7269\u4ef6\u5075\u6e2c\u6a21\u578b\u7684\u7b2c\u4e00\u9078\u64c7<\/strong>\u3002<\/p>\n\n\n\n

2020 \u5e74\u7531\u65bc YOLOv1-v3 \u7684\u539f\u59cb\u4f5c\u8005\u6c7a\u5b9a\u96e2\u958b\u96fb\u8166\u8996\u89ba\u7684\u7814\u7a76\uff0c\u4e2d\u7814\u9662\u7684\u738b\u5efa\u582f\u535a\u58eb\u63a5\u624b\u4e86\u8a72\u7814\u7a76\u4e26\u63a8\u51fa YOLOv4\uff0c\u5728\u7576\u5e74\u662f\u901f\u5ea6\u8207\u8fa8\u8b58\u7387\u5747\u8861\u7684\u6700\u4f73\u6a21\u578b\uff0c\u5f8c\u4f86\u4ed6\u9084\u63a8\u51fa\u4e86 Scaled-YOLOv4\u3001YOLOR\u3001YOLOv7\u3001YOLOv9 \u7b49\u5f8c\u7e8c\u7248\u672c\uff0c\u672c\u6587\u5c07\u4ecb\u7d39<\/strong> YOLOv7 <\/strong>\u5230<\/strong> YOLOv10 <\/strong>\u7684\u6700\u65b0\u767c\u5c55\uff0c\u4ee5\u53ca\u63a2\u7d22\u4e00\u500b\u6700\u91cd\u8981\u7684\u554f\u984c\uff0c\u90a3\u5c31\u662f\u5728<\/strong> YOLOv10 <\/strong>\u63a8\u51fa\u5f8c\uff0c\u5c31\u7269\u4ef6\u5075\u6e2c\u61c9\u7528\u7684\u89d2\u5ea6\uff0c\u662f\u5426\u9084\u6709\u4f7f\u7528\u904e\u53bb\u7248\u672c\u7684\u9700\u6c42<\/strong>\u3002<\/p>\n\n\n\n

\u5728\u6b63\u5f0f\u9032\u5165\u5167\u6587\u524d\uff0c\u60f3\u8207\u5927\u5bb6\u5206\u4eab\u4e00\u500b\u597d\u6d88\u606f\uff0c\u6211\u5728 2022 \u5e74\u958b\u8a2d\u7684\u6df1\u5ea6\u5b78\u7fd2\u5f71\u50cf\u8fa8\u8b58 | YOLO\u7269\u4ef6\u5075\u6e2c\u5168\u89e3\u6790\u8ab2\u7a0b\uff0c\u5728 2024 \u5e74\u5e95\u65b0\u589e\u4e86 YOLOv10 \u7684\u5167\u5bb9\uff0c\u672a\u4f86\u4e5f\u5c07\u6301\u7e8c\u5730\u66f4\u65b0\u6559\u6750\u3002\u5982\u679c\u4f60\u60f3\u9032\u4e00\u6b65\u4e86\u89e3 YOLO \u7269\u4ef6\u5075\u6e2c\u7684\u61c9\u7528\uff0c\u6b61\u8fce\u9ede\u64ca \u9019\u88e1<\/a> \u67e5\u770b\u8ab2\u7a0b\u8a73\u60c5\u3002<\/strong><\/p>\n\n\n\n

YOLOv7 <\/strong><\/mark> : <\/strong>\u6b64\u6a21\u578b\u662f\u5728 2022 \u5e74\u63a8\u51fa\u7684\uff0c\u5176\u4e3b\u8981\u7684\u7279\u9ede\u662f\u63d0\u51fa\u4e86 ELAN\uff08Efficient Layer Aggregation Network\uff09\u7279\u5fb5\u8403\u53d6\u7684\u6a21\u7d44\u8207\u53c3\u6578\u91cd\u6574\u5316\uff08reparameterization\uff09\uff0cELAN\u57fa\u65bc\u738b\u535a\u58eb\u904e\u53bb\u6240\u63d0\u51fa\u7684 CSPNet\uff08Cross Stage Partial Network\uff09\uff0c\u5176\u80fd\u8f03\u4f73\u7684\u878d\u5408\u7db2\u8def\u6a21\u7d44\u4e2d\u7684\u7279\u5fb5\u3002\u53c3\u6578\u91cd\u6574\u5316\u88ab\u63d0\u51fa\u7684\u4e3b\u8981\u69cb\u60f3\u4f86\u81ea\u65bc\u6709\u4eba\u767c\u73fe 3×3 convolution \u662f CUDA \u9ad8\u5ea6\u6700\u4f73\u5316\u7684\u6a21\u7d44\uff0c\u4f46\u57fa\u65bc ResNet \u7684 Skip connection \u537b\u53c8\u5c0d\u6548\u80fd\u6709\u6975\u5927\u7684\u63d0\u5347\uff0c\u7136\u800c\uff0c\u904e\u53bb\u7814\u7a76\u4eba\u54e1\u767c\u73fe skip connection \u662f\u53ef\u4ee5\u8f49\u63db\u6210 3*3 convolution \u7684\uff0c\u56e0\u6b64\u6211\u5011\u53ef\u4ee5\u5728\u8a13\u7df4\u6642\u4f7f\u7528 ResNet \u7684 skip connection\uff0c\u4f46\u5728\u63a8\u8ad6\u6642\u8f49\u63db\u6210 3*3 convolution \u4f86\u52a0\u5feb\u63a8\u8ad6\u7684\u901f\u5ea6\u3002YOLOv7 \u5728\u7269\u4ef6\u5075\u6e2c\u6a21\u578b\u4e2d\uff0c\u865f\u7a31\u662f 5 FPS to 160 FPS \u4e2d\uff0c\u6709\u8457\u6700\u5feb\u7684\u901f\u5ea6\u4ee5\u53ca\u6700\u9ad8\u7684\u8fa8\u8b58\u7387\uff0c\u5982\u4e0b\u5716\u6240\u793a\uff0c YOLOv7 \u7684\u6240\u6709\u6b21\u7248\u672c\uff0c\u76f8\u8f03\u65bc\u5176\u5b83\u6a21\u578b\uff0c\u662f\u6700\u5de6\u4e0a\u65b9\uff0c\u800c\u8fa8\u8b58\u7387\u6700\u9ad8\u7684\u7248\u672c\u53ef\u4ee5\u9054\u5230 56.8% \u7684\u8fa8\u8b58\u7387\u3002<\/p>\n\n\n\n

\"\"<\/a><\/figure>\n\n\n\n

YOLOv8 <\/strong><\/mark> : <\/strong>\u8fd1\u5e74\u4f86\u5728\u6b50\u6d32\uff0c\u6709\u4e00\u5bb6\u516c\u53f8 Ultralytics \u6301\u7e8c\u7684\u63a8\u51fa\u4e00\u7cfb\u5217\u539f\u751f\u4f7f\u7528PyTorch \u958b\u767c\u7684 YOLO\uff08YOLOv1-v4 \u90fd\u662f\u4f7f\u7528\u958b\u767c\u96e3\u5ea6\u8f03\u5927\uff0c\u5176\u65bc C \u8a9e\u8a00\u7684Darknet\uff09\uff0c\u5f9e 2018 \u5e74\u7684 YOLOv3\uff0c\u8a72\u516c\u53f8\u6240\u958b\u6e90\u7684 YOLO \u4e00\u76f4\u5ee3\u53d7\u597d\u8a55\uff0c\u5f8c\u7e8c\u8a72\u516c\u53f8\u4e5f\u63a8\u51fa\u4e86 YOLOv5 \u4ee5\u53ca YOLOv8\uff0c YOLOv8 \u7684\u4e00\u500b\u6700\u4e3b\u8981\u7684\u7a81\u7834\u662f\uff0c\u5b83\u5c07\u7269\u4ef6\u7684 bounding box \u4ee5\u53ca\u6a5f\u7387\u7684\u65b9\u5f0f\u4f86\u63cf\u8ff0\uff0c\u4e3b\u8981\u7684\u89c0\u5bdf\u662f\u4e0b\u5716\u4e2d\u4f8b\u5982\u4e0b\u5716\u4e2d\uff0c\u6ed1\u6c34\u51b2\u6d6a\u677f\u7684\u5de6\u908a\u4ee5\u53ca\u5927\u8c61\u7684\u53f3\u908a\u4f4d\u7f6e\u4e26\u4e0d\u660e\u78ba\uff0c\u5982\u679c\u6211\u5011\u4f7f\u7528\u6a5f\u7387\u7684\u65b9\u5f0f\u4f86\u63cf\u8ff0\uff0c\u5c31\u6709\u6a5f\u6703\u5f97\u5230\u8f03\u4f73\u7684\u7d50\u679c\uff0c\u4e5f\u5c31\u662f\u6211\u5011\u5c07\u7269\u4ef6\u5075\u6e2c\u8f49\u63db\u6210\u4f30\u6e2c uncertainty \u7684\u554f\u984c\u3002<\/p>\n\n\n\n

\"\"<\/a><\/figure>\n\n\n\n

YOLOv8 \u7684\u7279\u9ede\u5728\u65bc\uff0c\u5728 640 \u89e3\u6790\u5ea6\u7684\u5f71\u50cf\u4e2d\uff0c\u76f8\u8f03\u65bc\u5176\u5b83\u7684\u6a21\u578b\uff0c\u5b83\u53ef\u4ee5\u5f97\u5230\u7565\u9ad8\u7684\u8fa8\u8b58\u7387\uff0c\u4ee5\u53ca\u7565\u5feb\u7684\u901f\u5ea6\uff0c\u5982\u4e0b\u5716\u6240\u793a\uff0c\u56e0\u6b64\uff0c\u5982\u679c\u662f\u8003\u91cf\u5230\u4f4e\u89e3\u6790\u5ea6\u5f71\u50cf\u7684\u61c9\u7528\uff0cYOLOv8 \u662f\u4e00\u500b\u8f03\u4f73\u7684\u9078\u64c7\uff0c\u800c\u4e14\u6700\u91cd\u8981\u7684\u662f\uff0cUltralytics \u6b64\u516c\u53f8\u5c0d\u65bc\u6b64\u958b\u6e90\u5c08\u6848\u6301\u7e8c\u7dad\u8b77\uff0c\u5728\u8a72 repo\uff0c\u53ea\u8981\u4e0b\u4e0d\u540c\u7684\u6307\u4ee4\uff0c\u5c31\u53ef\u4ee5\u8f15\u6613\u4e32\u63a5\u4e0d\u540c\u5f71\u50cf\u4f86\u6e90\uff0c\u4f8b\u5982\u975c\u614b\u5716\u6a94\u6216\u662f\u7db2\u8def\u651d\u5f71\u6a5f\uff0c\u9084\u53ef\u4ee5\u5c07\u6a21\u578b\u8f49\u6210\u5728 NVIDIA \u5d4c\u5165\u5f0f\u5e73\u53f0\u4e0a\uff0c\u53ef\u4ee5\u6709\u8f03\u5feb\u63a8\u8ad6\u901f\u5ea6\u7684 TensorRT \u7248\u6a21\u578b\u3002<\/p>\n\n\n\n

\"\"<\/a><\/figure>\n\n\n\n

YOLOv9 <\/strong><\/mark> : <\/strong>\u4e2d\u7814\u9662\u7684\u738b\u535a\u58eb\u5728 2024 \u5e74\u521d\u63a8\u51fa\u4e86 YOLOv9\uff0c\u5176\u61c9\u7528\u4e86PGI\uff08Programmable Gradient Information\uff09\u7684\u6982\u5ff5\u4f86\u63d0\u5347 YOLOv9 \u7684\u6548\u80fd\uff0cPGI \u6700\u4e3b\u8981\u7684\u7cbe\u795e\u5728\u65bc\uff0c\u5b83\u9032\u4e00\u6b65\u7684\u5728\u8a13\u7df4\u6642\u4f7f\u7528\u8f14\u52a9\u7d50\u69cb\uff0c\u5e6b\u52a9\u6a21\u578b\u8403\u53d6\u51fa\u66f4\u5177\u4ee3\u8868\u6027\u7684\u7279\u5fb5\uff0c\u800c\u8a72\u8f14\u52a9\u7d50\u69cb\u5728\u63a8\u8ad6\u6642\u4e26\u4e0d\u9700\u8981\uff0c\u56e0\u70ba YOLOv9 \u5728 640 \u89e3\u6790\u5ea6\u4e0b\uff0c\u5982\u4e0b\u5716\u6240\u793a\uff0c\u53ef\u4ee5\u9054\u5230\u63a5\u8fd1 56% \u7684\u8fa8\u8b58\u7387\uff0c\u7136\u800c\uff0c\u4e0b\u5716\u4e2d\u7684\u6a6b\u8ef8\u662f\u53c3\u6578\u7684\u6578\u91cf\u800c\u4e0d\u662f\u901f\u5ea6\uff0c\u5728\u5f8c\u7e8c\u6211\u500b\u4eba\u7684\u5be6\u9a57\u4e2d\u767c\u73fe\uff0cYOLOv9 \u5728\u8003\u91cf\u5230\u63a8\u8ad6\u901f\u5ea6\u4ee5\u53ca\u76f8\u540c\u8fa8\u8b58\u7387\u4e0a\u4e26\u4e0d\u662f\u6700\u5feb\u7684\uff0c\u5728\u6b64\u6211\u5011\u8981\u5f37\u8abf\uff0c\u53c3\u6578\u8f03\u5c11\u4e0d\u4ee3\u8868\u901f\u5ea6\u53ef\u4ee5\u66f4\u5feb\uff0c\u4f46\u9019\u53ea\u662f\u5728 NVIDIA \u7684 GPU \u4e0a\u5206\u6790\u7684\u7d50\u679c\uff0c\u5982\u679c\u6211\u5011\u6709\u7279\u898f\u786c\u9ad4\uff0c\u91dd\u5c0d\u6b64\u6a21\u578b\u9032\u884c\u6700\u4f73\u5316\uff0c\u6216\u8a31\u6709\u6a5f\u6703\u5f97\u5230\u66f4\u5feb\u7684\u901f\u5ea6\u3002<\/p>\n\n\n\n

\"\"<\/a><\/figure>\n\n\n\n

YOLOv10 <\/strong><\/mark> <\/strong>: <\/strong>\u4f86\u81ea\u5c0d\u5cb8\u6e05\u83ef\u5927\u5b78\u7684\u5718\u968a\u5728 2024 \u5e74\u4e2d\uff0c\u518d\u6b21\u63a8\u51fa\u4e86 YOLOv10 \uff0c\u5176\u6700\u4e3b\u8981\u7684\u89c0\u5bdf\u5c31\u662f\u904e\u53bb YOLO \u7684\u6240\u6709\u7248\u672c\uff0c\u90fd\u6c92\u6709\u7d0d\u5165 NMS\uff08Non-Maximal-Suppression\uff09\u7684\u5206\u6790\uff0c\u4e5f\u5c31\u662f\u904e\u53bb\u8a55\u4f30\u63a8\u8ad6\u901f\u5ea6\u90fd\u6c92\u7d0d\u5165 NMS\u3002\u8a72\u5718\u968a\u70ba YOLOv10 \u8a2d\u8a08\u4e86\u5169\u500b Head\uff0c\u5206\u5225\u662f One-to-Many \u8207 One-to-One\uff0c\u524d\u8005\u7684\u610f\u601d\u662f\u540c\u4e00\u500b\u7269\u9ad4\u4e88\u8a31\u6a21\u578b\u8f38\u51fa\u8907\u6578\u500b\u5075\u6e2c\u7d50\u679c\uff0c\u5f8c\u8005\u53ea\u4e88\u8a31\u4e00\u500b\uff0c\u4e26\u671f\u5f85\u5169\u500b Head \u7684\u9810\u6e2c\u7d50\u679c\u80fd\u4e00\u81f4\uff0c\u4e5f\u8a31\u662f\u8aaa\uff0c\u5982\u679c\u6a21\u578b\u53ea\u8f38\u51fa\u4e00\u500b\u9810\u6e2c\u7d50\u679c\uff0c\u90a3 NMS \u5c31\u4e0d\u9700\u8981\u4e86\u3002\u5982\u4e0b\u5716\u6240\u793a\uff0cYOLOv10 \u53ef\u4ee5\u76f4\u63a5\u8f38\u51fa\u53f3\u65b9\u7684\u7d50\u679c\uff0c\u800c\u4e0d\u9700\u8981\u5148\u5f97\u5230\u5de6\u908a\uff0c\u91dd\u5c0d\u540c\u4e00\u53f0\u8eca\u5f97\u5230\u8907\u6578\u500b\u5075\u6e2c\u7d50\u679c\uff0c\u518d\u4f7f\u7528 NMS \u4f86\u6c7a\u5b9a\u552f\u4e00\u7684\u8f38\u51fa\u3002<\/p>\n\n\n\n

\"\"<\/a><\/figure>\n\n\n\n

\u5728 YOLOv10 \u7684\u5be6\u9a57\u5716\u8868\u4e2d\uff0c\u6211\u5011\u53ef\u4ee5\u770b\u5230\u5176\u5b83\u6a21\u578b Latency \u8207 Latencyf <\/sup>\u5c31 ms \u4f86\u8aaa\u6709\u5de8\u5927\u7684\u5dee\u5225\uff0c\u4f8b\u5982 YOLOv8-N\uff0c\u5982\u679c\u6c92\u6709 NMS\uff0c\u901f\u5ea6\u6703\u5f9e 6.16ms \u964d\u5230 1.79ms \uff0c\u4f46 YOLOv10-N \u4f7f\u7528 One-to-One \u4ee5\u53ca One-to-Many \u9019\u5169\u500b head \u5075\u6e2c\u7269\u9ad4\uff0c\u5176\u8fa8\u8b58\u7387\u8207 Latency \u53ef\u4ee5\u9054\u5230 38.5AP\/1.79ms \u4ee5\u53ca 39.5AP\/1.84ms\uff0c\u4e5f\u5c31\u662f\u8aaa\uff0c\u4f7f\u7528 One-to-Many \u5176\u4ecd\u9700\u8981\u4f7f\u7528 NMS\uff0c\u6240\u4ee5\u901f\u5ea6\u7565\u6162\uff0c\u4f46\u4f7f\u7528 one-to-one \u7684 head\uff0clatency \u8207\u8fa8\u8b58\u7387\u90fd\u7565\u4f4e\uff0c\u6700\u91cd\u8981\u7684\u662f\uff0c\u5982\u679c\u628a NMS \u7684\u901f\u5ea6\u4e5f\u7d0d\u5165\u8003\u91cf\uff0cYOLOv10 \u6b64\u6a21\u578b\u7684\u901f\u5ea6\u9060\u5927\u65bc\u904e\u53bb\u6240\u6709\u6a21\u578b\uff0c\u96d6\u7136 YOLOv10 \u76f8\u8f03\u65bc\u5176\u5b83\u7248\u672c\uff0c\u5c31\u8fa8\u8b58\u7387\u7684\u6bd4\u8f03\u7565\u986f\u4e0d\u516c\uff0c\u4f46\u8003\u91cf\u5230\u6574\u9ad4\u901f\u5ea6\uff0cYOLOv10 \u4ecd\u7136\u662f\u5d4c\u5165\u5f0f\u7cfb\u7d71\u6216\u662f\u908a\u7de3\u8a2d\u5099\u4e0a\u7684\u6700\u4f73\u6a21\u578b<\/strong>\u3002<\/p>\n\n\n\n

\"\"<\/a><\/figure>\n\n\n\n

\u91dd\u5c0d\u5404\u7a2e\u61c9\u7528\uff0c\u5728 2024 \u5e74\u5e95\u7684\u73fe\u5728\uff0cYOLOv7-YOLOv10 \u4ecd\u6709\u61c9\u7528\u7684\u6a5f\u6703\uff0c\u800c\u4e0d\u662f\u6700\u65b0\u4e00\u4ee3\u63a8\u51fa\u5f8c\uff0c\u904e\u53bb\u7684 YOLO \u7248\u672c\u5c31\u5b8c\u5168\u904e\u6642\u4e86\uff0c\u5728\u6b64\uff0c\u6211\u5011\u7e3d\u7d50\u9019\u56db\u4ee3 YOLO \u7684\u4f7f\u7528\u60c5\u5883\u3002<\/p>\n\n\n\n

\u5982\u679c\u4f60\u7684\u61c9\u7528\u9700\u8981\u4f7f\u7528\u5927\u89e3\u6790\u5ea6\u7684\u5f71\u50cf\uff081280\uff09\uff0c\u4f60\u53ef\u4ee5\u4f7f\u7528 YOLOv7\uff0c\u5982\u679c\u4f60\u7684\u786c\u9ad4\u6700\u9ad8\u53ea\u80fd\u63a5\u53d7 640 \u89e3\u6790\u5ea6\u7684\u5f71\u50cf\u53c8\u9700\u8981\u8f03\u5feb\u7684\u901f\u5ea6\uff0c\u8acb\u4f7f\u7528 YOLOv8\uff0c\u5982\u679c\u4f60\u8ffd\u6c42 640 \u89e3\u6790\u5ea6\u4e0b\u7684\u6700\u9ad8\u8fa8\u8b58\u7387\uff0c\u8acb\u4f7f\u7528 YOLOv9\uff0c\u5982\u679c\u5728\u4f60\u7684\u61c9\u7528\u4e2d\uff0c\u7269\u9ad4\u6642\u5e38\u91cd\u758a\uff0c\u8acb\u4f7f\u7528 YOLOv10\uff0c\u56e0\u70ba\u5b83\u53ef\u4ee5\u5728\u4e0d\u9700\u8981 NMS\u4e0b\u5f97\u5230\u66f4\u597d\u7684\u7d50\u679c\u3002<\/strong><\/p>\n\n\n\n

\u5f9e 2022 \u5e74\uff0c\u6211\u5728 TibaMe \u63a8\u51fa \u6df1\u5ea6\u5b78\u7fd2\u5f71\u50cf\u8fa8\u8b58 | YOLO\u7269\u4ef6\u5075\u6e2c\u5168\u89e3\u6790<\/a> \u8ab2\u7a0b\u4ee5\u4f86\uff0c\u6211\u4ecd\u4e0d\u65b7\u5730\u66f4\u65b0\u6559\u6750\uff0c\u5728 2024 \u5e74\u5e95\uff0c\u82e5\u4f60\u5e0c\u671b\u66f4\u6df1\u5165\u4e86\u89e3\u7269\u4ef6\u5075\u6e2c\u7684\u5be6\u969b\u61c9\u7528\uff0c\u6b61\u8fce\u4f60\u52a0\u5165\u8ab2\u7a0b\uff0c\u8207\u6211\u5011\u4e00\u8d77\u5728 YOLO \u7684\u4e16\u754c\u88e1\uff0c\u4e00\u8d77\u63a2\u7d22\u7269\u4ef6\u5075\u6e2c\u6a21\u578b\u7684\u6975\u9650\uff0c\u6210\u70ba YOLO \u7269\u4ef6\u5075\u6e2c\u7684\u5c08\u5bb6 !<\/p>\n\n\n\n

\"\"<\/a><\/figure>\n\n\n\n

\u4f60\u5c07\u80fd\u5728\u9019\u5802\u8ab2\u7a0b\u4e2d\u5b78\u7fd2\u5230\u2026<\/strong><\/p>\n\n\n\n