You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
End-to-end Self-Supervised Visual Feature Clustering project. Extracts deep representations using frozen ResNet50 transfer features from CIFAR-10 data, benchmarking K-Means, DBSCAN, and GMM. Features an interactive UMAP web dashboard and real-time visual query search engine.
Educational implementation of V-JEPA (Video Joint Embedding Predictive Architecture) in PyTorch. Includes Conv3D tubelet embeddings, Video ViT, EMA target encoder, spatiotemporal masking, linear probing, retrieval evaluation, and effective-rank analysis on Something-Something V2.