Jun 6 17:41 Sync from GitHub via hub-sync Mar 31 09:38 Add segment-objects.py for pixel-level image segmentation davanstrien +1: Claude Opus 4.6 (1M context) HF Staff New script that produces segmentation masks (semantic maps or per-instance
binary masks) using SAM3 with text prompts. Tested on HF Jobs with wildlife
camera trap images. Also updates README to document both scripts and adds
example segmentation image.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Feb 1 10:55 Add hf-jobs tag to README frontmatter davanstrien +1: Claude Opus 4.5 HF Staff Standardizing metadata tags across uv-scripts organization
for better discoverability of HF Jobs-compatible scripts.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> Nov 21, 2025 14:43 Update README.md 11:53 Improve formatting in README for better readability and consistency 11:47 Fix formatting in detection output message for better readability 11:43 Update example detection image in README for better responsiveness and visibility 11:43 Increase height of example detection image in README for improved visibility 11:41 Reduce height of example detection image in README for better display 11:41 Update example detection image in README to use HTML for better display 11:40 Update example detection image in README to use HTML for better display 11:38 Add example output section and image to README for object detection results 11:23 Add README.md for SAM3 object detection with usage instructions and examples 11:22 Add dataset card creation and visualization script for object detection results Nov 20, 2025 20:58 Convert boxes and scores to float32 before numpy conversion in process_batch for compatibility 20:51 Refactor imports for better organization and enable high-performance mode 20:48 Refactor process_batch to return detections in list format for improved consistency 20:42 Refactor output format to use dict-of-lists for object detections and update output schema accordingly 20:35 Uncomment features parameter in object detection function call 20:35 Comment out features parameter in object detection function call 20:31 Refactor output schema to use list for objects and simplify feature definitions 19:48 Refactor output schema creation to define features before processing dataset 19:46 Enhance batch processing to set tensor dtype for model compatibility 19:33 Refactor image processing in batch to handle grayscale images and unify prompts for model input 19:24 Refactor object detection script to use single class detection and update argument naming 19:09 Enhance input processing by moving tensors to the model's device and converting to the appropriate dtype 19:07 Refactor 19:04 Comment out torchvision and accelerate dependencies in detect-objects.py 18:55 draft] 17:50 initial commit 
