Official PyTorch implementation of the paper "Dataset Distillation with Neural Characteristic Function: A Minmax Perspective" (NCFM) in CVPR 2025 (Full Score, Highlight).

MergeIT: From Selection to Merging for Efficient Instruction Tuning

Data distillation benchmark

[IEEE TIP 2024] Normalizing Batch Normalization for Long-Tailed Recognition

[RSS2024] A Multi-Modal Large Language Model with Retrieval-augmented In-context Learning capacity designed for generalisable and explainable end-to-end driving

OmniGen: Unified Image Generation. https://arxiv.org/pdf/2409.11340

[NeurIPS 2024] Touchstone - Benchmarking AI on 5,172 o.o.d. CT volumes and 9 anatomical structures

🔥🔥First-ever hour scale video understanding models

COALA: A Practical and Vision-Centric Federated Learning Platform, accepted to ICML'24

The official repo for "SpatialBot: Precise Spatial Understanding with Vision Language Models.

🔥🔥MLVU: Multi-task Long Video Understanding Benchmark

M3D: Advancing 3D Medical Image Analysis with Multi-Modal Large Language Models

We introduce a novel approach for parameter generation, named neural network parameter diffusion (p-diff), which employs a standard latent diffusion model to synthesize a new set of parameters

[NeurIPS 2023] AbdomenAtlas 1.0 (5,195 CT volumes + 9 annotated classes)

A family of lightweight multimodal models.

Unofficial pytorch implementation of DDVM.

The official code for "SegVol: Universal and Interactive Volumetric Medical Image Segmentation".

SVIT: Scaling up Visual Instruction Tuning

Dataset pruning for ImageNet and LAION-2B.

[ICLR 2024] Real-Fake: Effective Training Data Synthesis Through Distribution Matching

A collection of visual instruction tuning datasets.