ResNet FPN Object Detection
Built a Feature Pyramid Network with a ResNet-50 backbone with a custom Torch loss function for rotated bounding boxes. Used it for object detection and segmentation on adversarial noisy binary images of stars, achieving over 96% IOU accuracy.
Install / Use
npx skills add yashpansari/ResNet-FPN-Object-DetectionInstalls into whichever agent you are using.
README
ResNet-FPN-Object-Detection
Built a Feature Pyramid Network with a ResNet-50 backbone with a custom Torch loss function for rotated bounding boxes. Used it for object detection and segmentation on adversarial noisy binary images of stars, achieving over 96% IOU accuracy wit under 1.5 Million parameters. Trained on 1 standard Tesla GPU for 5 hours (150 epochs).
Papers cited within code.
The data synthesizer generates images and labels. The model determines whether each image has a star and, if so, finds a rotated bounding box that bounds the star.
More precisely, the labels contain the following five numbers, which the model predicts:
- the x and y coordinates of the center
- yaw
- width and height.
The model is evaluated quantitatively using compute_score.py. The metric is the percent of correctly identified stars based on an IOU (Intersection of Union) threshold of 0.7 (for 1024 random samples).
Related Skills
node-connect
385.5kDiagnose OpenClaw Android, iOS, or macOS node pairing, QR/setup code, route, auth, and connection failures.
blender-python-addon
40.5kBlender Python add-on rules for operators, panels, properties, registration, testing, and API-safe scripting
flutter-development-guidelines-cursorrules-prompt-file
40.5kCursor rules for Flutter development with MVVM architecture, Riverpod state management, Material widgets, and Dart style guidelines.
commit-push-pr
140.7kCommit, push, and open a PR
