Upload README.md with huggingface_hub

README.md

@@ -0,0 +1,40 @@
+# OpenThermalPose and OpenThermalPose2: Datasets of Annotated Thermal Human Poses
+
+This dataset card describes OpenThermalPose and its extension, OpenThermalPose2, two publicly available datasets containing annotated thermal images of human poses.  These datasets are valuable resources for research in thermal-based human pose estimation.
+
+## OpenThermalPose
+
+OpenThermalPose provides 6,090 images featuring 31 subjects and 14,315 annotated human instances.  Each instance includes a bounding box and 17 anatomical keypoints, consistent with the MS COCO Keypoint dataset's conventions.  The dataset captures diverse scenarios, encompassing various fitness exercises, multiple-person activities, and outdoor walking under varied weather conditions.
+
+**Dataset Download:** [https://drive.google.com/file/d/1C5ThcFZm1twYtEta8GWUe1SENc9ER_0t/view?usp=sharing](https://drive.google.com/file/d/1C5ThcFZm1twYtEta8GWUe1SENc9ER_0t/view?usp=sharing)
+
+**Pre-trained YOLOv8-pose Models:** [https://drive.google.com/file/d/1BS2AB6wGRjZ8Tvz44_jkURR6moRpVS1j/view?usp=sharing](https://drive.google.com/file/d/1BS2AB6wGRjZ8Tvz44_jkURR6moRpVS1j/view?usp=sharing)
+
+
+## OpenThermalPose2
+
+OpenThermalPose2 expands upon OpenThermalPose, offering a significantly larger dataset with 11,391 images, 170 subjects, and 21,125 annotated human instances.  It maintains the annotation format of OpenThermalPose and adds scenarios including persons sitting in indoor environments.  The dataset covers various fitness exercises, multiple-person activities, and outdoor walking under varied weather conditions. YOLOv8-pose and YOLO11-pose models were trained and evaluated on this dataset.
+
+**Dataset Download:** [https://drive.google.com/file/d/1BDVprz9NtenCp3wDovA2lfKVBzJDCTt2/view?usp=sharing](https://drive.google.com/file/d/1BDVprz9NtenCp3wDovA2lfKVBzJDCTt2/view?usp=sharing)
+
+**Pre-trained YOLOv8-pose and YOLO11-pose Models:** [https://drive.google.com/file/d/19bvKSNKs3Z-8EFSJcTMaI-MdkNwVKh1f/view?usp=sharing](https://drive.google.com/file/d/19bvKSNKs3Z-8EFSJcTMaI-MdkNwVKh1f/view?usp=sharing)
+
+
+## Example Images
+
+**(Image examples illustrating sports exercises, outdoor walking, and indoor sitting scenarios are included here.  The original image URLs are retained.)**
+
+#### Sports exercises and two-person activities in an indoor environment
+<img src="https://github.com/IS2AI/OpenThermalPose/blob/main/session_1_2.png"> 
+
+#### Walking in outdoor environments
+<img src="https://github.com/IS2AI/OpenThermalPose/blob/main/3370.png"> 
+
+#### Sitting in an indoor environment
+<img src="https://github.com/IS2AI/OpenThermalPose/blob/main/102_1_2_4_15_101_1.png"> 
+
+##  Model Training and Evaluation
+
+Both YOLOv8-pose and YOLO11-pose models (nano, small, medium, large, and x-large) were trained and evaluated on these datasets using the Ultralytics library.  Further details on model training, validation, and prediction can be found in the Ultralytics documentation: [https://docs.ultralytics.com/tasks/pose/](https://docs.ultralytics.com/tasks/pose/)
+
+