support the newest ultralytics

Author: JackWoo0831

README.md

@@ -21,16 +21,15 @@ git checkout v2  # change to v2 branch !!
 
 ## 🗺️ Latest News
 
-- ***2024.11.29*** Fix bugs of C-BIoU Track (the state prediction and updating bugs)
-- ***2024.10.24*** Add Hybrid SORT and fix some errors and bugs of OC-SORT.
+- ***2025.4.3*** Support the newest ultralytics version (YOLO v3 ~ v12) and fix some bugs of hybrid sort.
 
 ## ❤️ Introduction
 
 This repo is a toolbox that implements the **tracking-by-detection paradigm multi-object tracker**. The detector supports:
 
 - YOLOX 
 - YOLO v7
-- YOLO v8, 
+- YOLO v3 ~ v12 by [ultralytics](https://docs.ultralytics.com/), 
 
 and the tracker supports:
 
@@ -61,7 +60,7 @@ The highlights are:
 ##  🔨 Installation
 
 The basic env is:
-- Ubuntu 18.04
+- Ubuntu 20.04
 - Python：3.9, Pytorch: 1.12
 
 Run following commond to install other packages:
@@ -86,12 +85,14 @@ python3 setup.py develop
 
 There is no need to execute addtional steps as the repo itself is based on YOLOv7.
 
-3. YOLO v8:
+3. YOLO series by ultralytics:
 
 Please run:
 
 ```bash
-pip3 install ultralytics==8.0.94
+pip3 install ultralytics
+or 
+pip3 install --upgrade ultralytics
 ```
 
 ### 📑 Data preparation
@@ -148,7 +149,11 @@ Some references may help you:
 python train_aux.py --dataset visdrone --workers 8 --device <$GPU_id$> --batch-size 16 --data data/visdrone_all.yaml --img 1280 1280 --cfg cfg/training/yolov7-w6.yaml --weights <$YOLO v7 pretrained model path$> --name yolov7-w6-custom --hyp data/hyp.scratch.custom.yaml
 ```  
 
-- YOLO v8: `tracker/yolov8_utils/train_yolov8.py`
+- YOLO series (YOLO v3 ~ v12) by ultralytics:: `tracker/yolo_ultralytics_utils/train_yolo_ultralytics.py`
+
+```shell
+python tracker/yolo_ultralytics_utils/train_yolo_ultralytics.py --model_weight weights/yolo11m.pt --data_cfg tracker/yolo_ultralytics_utils/data_cfgs/visdrone_det.yaml --epochs 30 --batch_size 8 --img_sz 1280 --device 0
+```
 
 
 
@@ -157,42 +162,47 @@ python train_aux.py --dataset visdrone --workers 8 --device <$GPU_id$> --batch-s
 If you only want to run a demo:
 
 ```bash
-python tracker/track_demo.py --obj ${video path or images folder path} --detector ${yolox, yolov8 or yolov7} --tracker ${tracker name} --kalman_format ${kalman format, sort, byte, ...} --detector_model_path ${detector weight path} --save_images
+python tracker/track_demo.py --obj ${video path or images folder path} --detector ${yolox, yolov7 or yolo_ultra} --tracker ${tracker name} --kalman_format ${kalman format, sort, byte, ...} --detector_model_path ${detector weight path} --save_images
 ```
 
+> ❗❗Important Notes
+> 
+> If you want to use the detector trained by **ultralytics**, the `--detector` argument **must include** the substring `ultra`, such as 
+> `--detector yolo_ultra`, `--detector yolo_ultra_v8`, `--detector yolov11_ultra`, `--detector yolo12_ultralytics`, etc.
+
 For example:
 
 ```bash
-python tracker/track_demo.py --obj M0203.mp4 --detector yolov8 --tracker deepsort --kalman_format byte --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt --save_images
+python tracker/track_demo.py --obj M0203.mp4 --detector yolo_ultra_v8 --tracker deepsort --kalman_format byte --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt --save_images
 ```
 
 If you want to run trackers on dataset:
 
 ```bash
-python tracker/track.py --dataset ${dataset name, related with the yaml file} --detector ${yolox, yolov8 or yolov7} --tracker ${tracker name} --kalman_format ${kalman format, sort, byte, ...} --detector_model_path ${detector weight path}
+python tracker/track.py --dataset ${dataset name, related with the yaml file} --detector ${yolox, yolo_ultra_v8 or yolov7} --tracker ${tracker name} --kalman_format ${kalman format, sort, byte, ...} --detector_model_path ${detector weight path}
 ```
 
 For example:
 
-- SORT: `python tracker/track.py --dataset uavdt --detector yolov8 --tracker sort --kalman_format sort --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt `
+- SORT: `python tracker/track.py --dataset uavdt --detector yolo_ultra_v8 --tracker sort --kalman_format sort --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt `
 
-- DeepSORT: `python tracker/track.py --dataset uavdt --detector yolov7 --tracker deepsort --kalman_format byte --detector_model_path weights/yolov7_UAVDT_35epochs_20230507.pt`
+- DeepSORT: `python tracker/track.py --dataset visdrone_part --detector yolov7 --tracker deepsort --kalman_format byte --detector_model_path weights/yolov8l_VisDroneDet_35epochs_20230605.pt`
 
-- ByteTrack: `python tracker/track.py --dataset uavdt --detector yolov8 --tracker bytetrack --kalman_format byte --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
+- ByteTrack: `python tracker/track.py --dataset uavdt --detector yolo_ultra_v8 --tracker bytetrack --kalman_format byte --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
 
 - OCSort: `python tracker/track.py --dataset mot17 --detector yolox --tracker ocsort --kalman_format ocsort --detector_model_path weights/bytetrack_m_mot17.pth.tar`
 
-- C-BIoU Track: `python tracker/track.py --dataset uavdt --detector yolov8 --tracker c_bioutrack --kalman_format bot --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
+- C-BIoU Track: `python tracker/track.py --dataset uavdt --detector yolo_ultra_v8 --tracker c_bioutrack --kalman_format bot --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
 
 - BoT-SORT: `python tracker/track.py --dataset uavdt --detector yolox --tracker botsort --kalman_format bot --detector_model_path weights/yolox_m_uavdt_50epochs.pth.tar`
 
-- Strong SORT: `python tracker/track.py --dataset uavdt --detector yolov8 --tracker strongsort --kalman_format strongsort --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
+- Strong SORT: `python tracker/track.py --dataset visdrone_part --detector yolo_ultra_v8 --tracker strongsort --kalman_format strongsort --detector_model_path weights/yolov8l_VisDrone_35epochs_20230509.pt`
 
-- Sparse Track: `python tracker/track.py --dataset uavdt --detector yolov8 --tracker sparsetrack --kalman_format bot --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
+- Sparse Track: `python tracker/track.py --dataset uavdt --detector yolo_ultra_v11 --tracker sparsetrack --kalman_format bot --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
 
 - UCMC Track: `python tracker/track.py --dataset mot17 --detector yolox --tracker ucmctrack --kalman_format ucmc --detector_model_path weights/bytetrack_m_mot17.pth.tar --camera_parameter_folder ./tracker/cam_param_files`
 
-- Hybrid SORT: `python tracker/track.py --dataset mot17 --detector yolox --tracker hybridsort --kalman_format hybridsort --detector_model_path weights/bytetrack_m_mot17.pth.tar --save_images`
+- Hybrid SORT: `python tracker/track.py --dataset visdrone_part --detector yolo_ultra --tracker hybridsort --kalman_format hybridsort --detector_model_path weights/yolov8l_VisDrone_35epochs_20230509.pt --save_images`
 
 > **Important notes for UCMC Track:**
 > 

README_CN.md

@@ -15,8 +15,7 @@ git checkout v2  # change to v2 branch !!
 
 ## 🗺️ 最近更新
 
-- ***2024.11.29*** 修复了C-BIoU Tracker中轨迹状态的更新和预测的错误
-- ***2024.10.24*** 增加了 Hybrid SORT 并且修复了OC-SORT的一些bug和错误。
+- ***2025.4.3*** 增加了ultralytics库最新版本的支持，修复了hybrid sort中的一些bug.
 
 
 ## ❤️ 介绍
@@ -25,7 +24,7 @@ git checkout v2  # change to v2 branch !!
 
 - YOLOX 
 - YOLO v7
-- YOLO v8, 
+- YOLO v3 ~ v12 by [ultralytics](https://docs.ultralytics.com/), 
 
 跟踪器支持:
 
@@ -60,7 +59,7 @@ REID模型支持：
 ##  🔨 安装
 
 基本环境是：
-- Ubuntu 18.04
+- Ubuntu 20.04
 - Python：3.9, Pytorch: 1.12
 
 运行以下命令安装其他包：
@@ -85,12 +84,14 @@ python3 setup.py develop
 
 由于仓库本身就是基于YOLOv7的，因此无需执行额外的步骤。
 
-3. YOLO v8:
+3. Ultralytics的YOLO系列模型：
 
 请运行：
 
 ```bash
-pip3 install ultralytics==8.0.94
+pip3 install ultralytics
+or 
+pip3 install --upgrade ultralytics
 ```
 
 ### 📑 数据准备
@@ -147,7 +148,11 @@ CATEGORY_DICT:
 python train_aux.py --dataset visdrone --workers 8 --device <$GPU_id$> --batch-size 16 --data data/visdrone_all.yaml --img 1280 1280 --cfg cfg/training/yolov7-w6.yaml --weights <$YOLO v7 pretrained model path$> --name yolov7-w6-custom --hyp data/hyp.scratch.custom.yaml
 ```  
 
-- YOLO v8: `tracker/yolov8_utils/train_yolov8.py`
+- Ultralytics的YOLO系列模型 (YOLO v3 ~ v12): `tracker/yolo_ultralytics_utils/train_yolo_ultralytics.py`
+
+```shell
+python tracker/yolo_ultralytics_utils/train_yolo_ultralytics.py --model_weight weights/yolo11m.pt --data_cfg tracker/yolo_ultralytics_utils/data_cfgs/visdrone_det.yaml --epochs 30 --batch_size 8 --img_sz 1280 --device 0
+```
 
 
 
@@ -156,9 +161,14 @@ python train_aux.py --dataset visdrone --workers 8 --device <$GPU_id$> --batch-s
 如果你只是想运行一个demo:
 
 ```bash
-python tracker/track_demo.py --obj ${video path or images folder path} --detector ${yolox, yolov8 or yolov7} --tracker ${tracker name} --kalman_format ${kalman format, sort, byte, ...} --detector_model_path ${detector weight path} --save_images
+python tracker/track_demo.py --obj ${video path or images folder path} --detector ${yolox, yolov7 or yolo_ultra} --tracker ${tracker name} --kalman_format ${kalman format, sort, byte, ...} --detector_model_path ${detector weight path} --save_images
 ```
 
+> ❗❗重要提示
+> 
+> 如果你是通过 **ultralytics** 库训练检测模型, 命令里的`--detector`参数 **必须包含**`ultra`字段, 例如
+> `--detector yolo_ultra`, `--detector yolo_ultra_v8`, `--detector yolov11_ultra`, `--detector yolo12_ultralytics`, 等等.
+
 例如:
 
 ```bash
@@ -168,26 +178,30 @@ python tracker/track_demo.py --obj M0203.mp4 --detector yolov8 --tracker deepsor
 如果你想在数据集上测试:
 
 ```bash
-python tracker/track.py --dataset ${dataset name, related with the yaml file} --detector ${yolox, yolov8 or yolov7} --tracker ${tracker name} --kalman_format ${kalman format, sort, byte, ...} --detector_model_path ${detector weight path}
+python tracker/track.py --dataset ${dataset name, related with the yaml file} --detector ${yolox, yolo_ultra_v8 or yolov7} --tracker ${tracker name} --kalman_format ${kalman format, sort, byte, ...} --detector_model_path ${detector weight path}
 ```
 
 例如:
 
-- SORT: `python tracker/track.py --dataset uavdt --detector yolov8 --tracker sort --kalman_format sort --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt `
+- SORT: `python tracker/track.py --dataset uavdt --detector yolo_ultra_v8 --tracker sort --kalman_format sort --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt `
 
-- DeepSORT: `python tracker/track.py --dataset uavdt --detector yolov7 --tracker deepsort --kalman_format byte --detector_model_path weights/yolov7_UAVDT_35epochs_20230507.pt`
+- DeepSORT: `python tracker/track.py --dataset visdrone_part --detector yolov7 --tracker deepsort --kalman_format byte --detector_model_path weights/yolov8l_VisDroneDet_35epochs_20230605.pt`
 
-- ByteTrack: `python tracker/track.py --dataset uavdt --detector yolov8 --tracker bytetrack --kalman_format byte --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
+- ByteTrack: `python tracker/track.py --dataset uavdt --detector yolo_ultra_v8 --tracker bytetrack --kalman_format byte --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
 
-- OCSort: `python tracker/track.py --dataset uavdt --detector yolov8 --tracker ocsort --kalman_format ocsort --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
+- OCSort: `python tracker/track.py --dataset mot17 --detector yolox --tracker ocsort --kalman_format ocsort --detector_model_path weights/bytetrack_m_mot17.pth.tar`
 
-- C-BIoU Track: `python tracker/track.py --dataset uavdt --detector yolov8 --tracker c_bioutrack --kalman_format bot --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
+- C-BIoU Track: `python tracker/track.py --dataset uavdt --detector yolo_ultra_v8 --tracker c_bioutrack --kalman_format bot --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
 
 - BoT-SORT: `python tracker/track.py --dataset uavdt --detector yolox --tracker botsort --kalman_format bot --detector_model_path weights/yolox_m_uavdt_50epochs.pth.tar`
 
+- Strong SORT: `python tracker/track.py --dataset visdrone_part --detector yolo_ultra_v8 --tracker strongsort --kalman_format strongsort --detector_model_path weights/yolov8l_VisDroneDet_35epochs_20230605.pt`
+
+- Sparse Track: `python tracker/track.py --dataset uavdt --detector yolo_ultra_v11 --tracker sparsetrack --kalman_format bot --detector_model_path weights/yolov8l_UAVDT_60epochs_20230509.pt`
+
 - UCMC Track: `python tracker/track.py --dataset mot17 --detector yolox --tracker ucmctrack --kalman_format ucmc --detector_model_path weights/bytetrack_m_mot17.pth.tar --camera_parameter_folder ./tracker/cam_param_files`
 
-- Hybrid SORT: `python tracker/track.py --dataset mot17 --detector yolox --tracker hybridsort --kalman_format hybridsort --detector_model_path weights/bytetrack_m_mot17.pth.tar --save_images`
+- Hybrid SORT: `python tracker/track.py --dataset visdrone_part --detector yolo_ultra --tracker hybridsort --kalman_format hybridsort --detector_model_path weights/yolov8l_VisDrone_35epochs_20230509.pt --save_images`
 
 >**UCMC Track的重要提示：**
 > 

tracker/config_files/visdrone_part.yaml

@@ -1,7 +1,7 @@
 # Config file of VisDrone dataset
 
 DATASET_ROOT: '/data/wujiapeng/datasets/VisDrone2019/VisDrone2019'
-SPLIT: test
+SPLIT: val
 CATEGORY_NAMES: 
   - 'pedestrain'
   - 'car'

tracker/track.py

@@ -54,14 +54,14 @@
     logger.warning('Load yolov7 fail. If you want to use yolov7, please check the installation.')
     pass
 
-# YOLOv8 modules
+# Ultralytics YOLO modules (support YOLOv3 ~ YOLOv12)
 try:
     from ultralytics import YOLO
-    from yolov8_utils.postprocess import postprocess as postprocess_yolov8
+    from yolo_ultralytics_utils.postprocess import postprocess as postprocess_ultralytics
 
 except Exception as e:
     logger.warning(e)
-    logger.warning('Load yolov8 fail. If you want to use yolov8, please check the installation.')
+    logger.warning('Load ultralytics fail. If you want to use ultralytics, please check the installation.')
     pass
 
 TRACKER_DICT = {
@@ -83,7 +83,7 @@ def get_args():
 
     """general"""
     parser.add_argument('--dataset', type=str, default='visdrone_part', help='visdrone, mot17, etc.')
-    parser.add_argument('--detector', type=str, default='yolov8', help='yolov7, yolox, etc.')
+    parser.add_argument('--detector', type=str, default='yolo_ultralytics_v8', help='yolov7, yolox, etc.')
     parser.add_argument('--tracker', type=str, default='sort', help='sort, deepsort, etc')
     parser.add_argument('--reid_model', type=str, default='osnet_x0_25', help='osnet or deppsort')
 
@@ -170,7 +170,7 @@ def main(args, dataset_cfgs):
 
         logger.info(f'Now detector is on device {next(model.parameters()).device}')
 
-    elif args.detector == 'yolov8':
+    elif 'ultra' in args.detector:
 
         logger.info(f"loading detector {args.detector} checkpoint {args.detector_model_path}")
         model = YOLO(args.detector_model_path)
@@ -182,6 +182,9 @@ def main(args, dataset_cfgs):
 
     else:
         logger.error(f"detector {args.detector} is not supprted")
+        logger.error("If you want to use the yolo v8 by ultralytics, please specify the `--detector` \
+                     as the string including the substring `ultra`, \
+                     such as `yolo_ultra_v8` or `yolo11_ultralytics`")
         exit(0)
 
     """3. load sequences"""
@@ -226,7 +229,7 @@ def main(args, dataset_cfgs):
             # start timing this frame
             timer.tic()
 
-            if args.detector == 'yolov8':
+            if 'ultra' in args.detector:
                 img = img.squeeze(0).cpu().numpy()
 
             else:
@@ -237,8 +240,8 @@ def main(args, dataset_cfgs):
 
             # get detector output 
             with torch.no_grad():
-                if args.detector == 'yolov8':
-                    output = model.predict(img, conf=args.conf_thresh, iou=args.nms_thresh)
+                if 'ultra' in args.detector:
+                    output = model.predict(img, conf=args.conf_thresh, iou=args.nms_thresh, verbose=False)
                 else:
                     output = model(img)
 
@@ -250,8 +253,8 @@ def main(args, dataset_cfgs):
             elif args.detector == 'yolov7':
                 output = postprocess_yolov7(output, args.conf_thresh, args.nms_thresh, img.shape[2:], ori_img.shape)
 
-            elif args.detector == 'yolov8':
-                output = postprocess_yolov8(output)
+            elif 'ultra' in args.detector:
+                output = postprocess_ultralytics(output)
 
             else: raise NotImplementedError
 

tracker/track_demo.py

@@ -53,10 +53,10 @@
     logger.warning('Load yolov7 fail. If you want to use yolov7, please check the installation.')
     pass
 
-# YOLOv8 modules
+# Ultralytics YOLO modules (support YOLOv3 ~ YOLOv12)
 try:
     from ultralytics import YOLO
-    from yolov8_utils.postprocess import postprocess as postprocess_yolov8
+    from yolo_ultralytics_utils.postprocess import postprocess as postprocess_ultralytics
 
 except Exception as e:
     logger.warning(e)
@@ -83,7 +83,7 @@ def get_args():
     """general"""
     parser.add_argument('--obj', type=str, required=True, default='demo.mp4', help='video or images folder PATH')
 
-    parser.add_argument('--detector', type=str, default='yolov8', help='yolov7, yolox, etc.')
+    parser.add_argument('--detector', type=str, default='yolo_ultralytics_v8', help='yolov7, yolox, etc.')
     parser.add_argument('--tracker', type=str, default='sort', help='sort, deepsort, etc')
     parser.add_argument('--reid_model', type=str, default='osnet_x0_25', help='osnet or deppsort')
 
@@ -168,7 +168,7 @@ def main(args):
 
         logger.info(f'Now detector is on device {next(model.parameters()).device}')
 
-    elif args.detector == 'yolov8':
+    elif 'ultra' in args.detector:
 
         logger.info(f"loading detector {args.detector} checkpoint {args.detector_model_path}")
         model = YOLO(args.detector_model_path)
@@ -180,6 +180,9 @@ def main(args):
 
     else:
         logger.error(f"detector {args.detector} is not supprted")
+        logger.error("If you want to use the yolo v8 by ultralytics, please specify the `--detector` \
+                as the string including the substring `ultra`, \
+                such as `yolo_ultra_v8` or `yolo11_ultralytics`")
         exit(0)
 
     """3. load sequences"""
@@ -200,7 +203,7 @@ def main(args):
     """4. Tracking"""
 
     for frame_idx, (ori_img, img) in process_bar:
-        if args.detector == 'yolov8':
+        if 'ultra' in args.detector:
             img = img.squeeze(0).cpu().numpy()
 
         else:
@@ -211,7 +214,7 @@ def main(args):
 
         # get detector output 
         with torch.no_grad():
-            if args.detector == 'yolov8':
+            if 'ultra' in args.detector:
                 output = model.predict(img, conf=args.conf_thresh, iou=args.nms_thresh)
             else:
                 output = model(img)
@@ -224,8 +227,8 @@ def main(args):
         elif args.detector == 'yolov7':
             output = postprocess_yolov7(output, args.conf_thresh, args.nms_thresh, img.shape[2:], ori_img.shape)
 
-        elif args.detector == 'yolov8':
-            output = postprocess_yolov8(output)
+        elif 'ultra' in args.detector:
+            output = postprocess_ultralytics(output)
 
         else: raise NotImplementedError