Xiaomi Research ORG

[ICLR 2026] ReCogDrive: A Reinforced Cognitive Framework for End-to-End Autonomous Driving

DGGT: Feedforward 4D Reconstruction of Dynamic Driving Scenes using Unposed Images

No description.

Efficient audio understanding with general audio captions

🤗 R1-AQA Model: mispeech/r1-aqa

UniDriveVLA: Unifying Understanding, Perception, and Action Planning for Autonomous Driving

ControlFoley: Unified and Controllable Video-to-Audio Generation with Cross-Modal Conflict Handling

No description.

No description.

SVOR - Stable Video Object Removal

[NeurIPS 2025] Think Silently, Think Fast: Dynamic Latent Compression of LLM Reasoning Chains

[NeurIPS'25] Time-R1: Post-Training Large Vision Language Model for Temporal Video Grounding

[NeurIPS 2025]Genesis: Multimodal Driving Scene Generation with Spatio-Temporal and Cross-Modal Consistency

Official code repository of Shuffle-R1

[ICCV 2025] Implementation of the paper "Q-Frame: Query-aware Frame Selection and Multi-Resolution Adaptation for Video-LLMs"

Official PyTorch inference code for the Interspeech 2025 paper: Efficient Speech Enhancement via Embeddings from Pre-trained Generative Audioencoders

Official Implementation of GLAP - General Language Audio Pretraining

[CVPR2026] DriveLaW: Unifying Planning and Video Generation in a Latent Driving World

XARES-LLM

[ICLR 2026] WorldSplat: Gaussian-Centric Feed-Forward 4D Scene Generation for Autonomous Driving

Gemma-based Multilingual Machine Translation Models

No description.

end-to-end text to audio scene generation model

A Xiaomi Research AI model that generates full audio scenes from text, blending spoken dialogue, background music, sound effects, and ambient sounds into a single WAV output file.

No description.

State-of-the-art continious audio tokenization

No description.

No description.

[CVPR'26] TimeViper: A Hybrid Mamba-Transformer Vision-Language Model for Efficient Long Video Understanding

GUIEvalKit: Open-source Evaluation Toolkit for GUI Agents

[NeurIPS 2025] Implementation of the paper "BTL-UI: Blink-Think-Link Reasoning Model for GUI Agent"

DAR introduces the diagonal scanning order for next-token prediction and proposes a direction-aware autoregressive transformer framework.

[NeurIPS 2025] Pixel-Perfect Depth

No description.

PROVE: A Perceptual RemOVal cohErence Benchmark for Visual Media

ProactiveMobile

No description.

No description.

DriveMRP: Enhancing Vision-Language Models with Synthetic Motion Data for Motion Risk Prediction

No description.

No description.

No description.

No description.

No description.

No description.

No description.

No description.