Highlights
- Pro
Stars
OpenFace – a state-of-the art tool intended for facial landmark detection, head pose estimation, facial action unit recognition, and eye-gaze estimation.
Generate diagrams from textual description
A Github repository about micro-expression recognition, micro-expression detection, and micro-expression analysis
Facial Emotion Recognition using OpenCV and Deepface
This is the official code of the paper: Facial Micro-motion-aware Mixup for Micro-expression Recognition
A Novel Micro-Expression Recognition Approach Using Attention-Based Magnification-Adaptive Networks, in ICASSP, 2022
[CVPR 2023] Micron-BERT: BERT-based Facial Micro-Expression Recognition
Scene-Centric Unsupervised Panoptic Segmentation (CVPR 2025 Highlight)
Visual Planning: Let's Think Only with Images
EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL
Project Page For "Seg-Zero: Reasoning-Chain Guided Segmentation via Cognitive Reinforcement"
This is the official implementation of the SEMamba paper. (Accepted to IEEE SLT 2024)
Official PyTorch implementation of the paper "Dataset Distillation with Neural Characteristic Function: A Minmax Perspective" (NCFM) in CVPR 2025 (Highlight).
[ECCV 2024] The official code of paper "Open-Vocabulary SAM".
⚡ TabPFN: Foundation Model for Tabular Data ⚡
Mulberry, an o1-like Reasoning and Reflection MLLM Implemented via Collective MCTS
LLaVA-CoT, a visual language model capable of spontaneous, systematic reasoning
The official implementation of the paper "LEGION: Learning to Ground and Explain for Synthetic Image Detection"
Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)
Microsoft's GraphRAG + AutoGen + Ollama + Chainlit = Fully Local & Free Multi-Agent RAG Superbot
Virtual whiteboard for sketching hand-drawn like diagrams
[ICLR 2025 Spotlight] The official implementation of the paper “LOKI：A Comprehensive Synthetic Data Detection Benchmark using Large Multimodal Models”
[CVPR 2025 Highlight] Official code for "Olympus: A Universal Task Router for Computer Vision Tasks"
Use PEFT or Full-parameter to CPT/SFT/DPO/GRPO 500+ LLMs (Qwen3, Qwen3-MoE, Llama4, InternLM3, DeepSeek-R1, ...) and 200+ MLLMs (Qwen2.5-VL, Qwen2.5-Omni, Qwen2-Audio, Ovis2, InternVL3, Llava, GLM4…