Our Research Featured ReMix: Training Generalized Person Re-identification on a Mixture of Data Our WACV 2025 paper demonstrates that intelligently leveraging a mixture of heterogeneous data can significantly enhance the quality of generalized person Re-ID.
Paper Review Paper Review: Matryoshka Representation Learning We can train the model with embeddings of different sizes without retraining.
Paper Review Featured Paper Review: TETRIS: Towards Exploring the Robustness of Interactive Segmentation It turns out that modern benchmarks incorrectly assess the robustness of interactive segmentation methods.
Paper Review How OpenAI's Sora works: key insights Work that will accelerate the development of text-to-video generative AI and create many problems with deepfakes.
Paper Review Paper Review: Weak-to-Strong Generalization by OpenAI OpenAI has made a significant step towards superintelligence, but there is still a long way to go.
Paper Review Paper Review: Real-World Humanoid Locomotion with Reinforcement Learning Robots have learned to decide for themselves what and how to do it.
