⬢github Python · 635 ★ +19 since we first saw it · pushed 1 d ago · MIT
PSRben/VisionHOPE
Official PyTorch implementation of VisionHOPE: Visual Backbones as Self-Modifying Learning Systems.
VisionHOPE is a PyTorch implementation of a computer vision backbone that treats visual processing as a self-modifying learning system. Built on self-referential nested learning (SRNL), it uses coupled memories that co-evolve with the learning rule while scanning an image. The repo includes ImageNet classification, COCO detection/segmentation, ADE20K segmentation tasks, standalone SRNL modules, fast inference, and pretrained checkpoints.
Why now: Fresh release of an arXiv preprint (late September 2026) with pretrained weights just published, quickly gaining stars as a new take on visual backbones beyond CNNs, ViTs, and SSMs.
Who it is for: Computer vision researchers and engineers looking for novel backbone architectures for classification, detection, and segmentation.
backbonebackbone-networkscomputer-visionimage-classificationvision-framework
Stars over our 12 snapshots: 616 to 635, since 2 h ago.
Where people talked about it
- ⬢github new repos, most starred 13 min ago
API: https://socialmediatrends-api.osmike.com/v1/repos/PSRben/VisionHOPE