2026.3.1.21.56

README.md

@@ -16,6 +16,14 @@ size_categories:
 ### A new benchmark dataset comprising diverse vehicle classes with frame-level tracking annotation introduced in paper: *FastTracker: Real-Time and Accurate Visual Tracking*
 _[Hamidreza Hashempoor](https://hamidreza-hashempoor.github.io/),  Yu Dong Hwang_.
 
+
+##  Updates
+
+| Date | Update |
+|------|---------|
+| **2026-03-01** |  Speciall thanks to _[Mikhail Kozak](https://huggingface.co/datasets/Fleyderer/FastTracker-Benchmark-MOT) which helped to prepare the revised current version of the benchmark.|
+
+
 ## Resources
 | Github  | Paper |
 |:-----------------:|:-------:|
@@ -71,20 +79,41 @@ Brief statistics and visualization of FastTracker benchmark and its comparison w
 
 ### Data Format
 
-GT format is like (each line):
-`frame, id, bb_left, bb_top, bb_width, bb_height, conf, class, 1.0`.
+Each sequence is provided as a compressed archive inside the train/ directory:
+```bash
+train/
+    task_day_left_turn.zip
+    task_day_occlusion.zip
+    ...
+```
 
-To prepare the dataset, first run `extract_frames.py` to decode frames from each video.  
-In **line 11** of the script, add the video filename and the number of frames you want to extract.  
+
+After extracting a sequence archive, the structure is:
 ```bash
-python extract_frames.py
+task_xxx/
+    img1/          # extracted frames (.jpg)
+    gt/            # MOT-format ground truth (gt.txt)
+    video/         # reconstructed video (.mp4) and metadata
+    seqinfo.ini    # sequence metadata
 ```
 
-Then, convert the ground truth into COCO format with:
+
+
+GT formats `gt/gt.txt` is like (each line):
+`frame, id, bb_left, bb_top, bb_width, bb_height, conf, class, visibility`.
+
+Each sequence includes `seqinfo.ini`:
 ```bash
-python convert_to_coco.py
+[Sequence]
+name=task_day_left_turn
+imDir=img1
+frameRate=30
+seqLength=1962
+imWidth=1920
+imHeight=1080
+imExt=.jpg
 ```
-This will generate annotations/train.json ready for training your detector.
+