An agentic skills framework & software development methodology that works.

Speaker Language Model

The official implementation of 'LAMB: LLM-based Audio Captioning with Modality Gap Bridging via Cauchy-Schwarz Divergence' (ICASSP2026)

LightlyStudio - The Unified Data Platform for Multimodal ML

[NeurIPS 2025] AVCD: Mitigating Hallucinations in Audio-Visual Large Language Models through Contrastive Decoding

Implementation of a single layer of the MMDiT, proposed in Stable Diffusion 3, in Pytorch

Interspeech_ablation_study (Boundary-Conscious Pruning: Hard Set-Aware Model Compression for Efficient Speaker Recognition)

Techniques and tools for optimizing how AI coding assistants understand your codebase, with a focus on cost reduction and efficiency.

[INTERSPEECH 2025] Official code for "SEED: Speaker Embedding Enhancement Diffusion Model"

[INTERSPEECH 2024] Official pytorch code for the paper "Disentangled Representation Learning for Environment-agnostic Speaker Recognition"

Official Pytorch Implementation of 'LAVCap: LLM-based Audio-Visual Captioning using Optimal Transport' (ICASSP2025)

Official Python SDK for the Agent2Agent (A2A) Protocol

Samples using the Agent2Agent (A2A) Protocol

Official code of Diffusion -bridge

A TTS model capable of generating ultra-realistic dialogue in one pass.

High accuracy RAG for answering questions from scientific documents with citations

Agent Laboratory is an end-to-end autonomous research workflow meant to assist you as the human researcher toward implementing your research ideas

Code and Pretrained Models for ICLR 2023 Paper "Contrastive Audio-Visual Masked Autoencoder".

Mind the Gap: Understanding the Modality Gap in Multi-modal Contrastive Representation Learning

Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities

CAIRI Supervised, Semi- and Self-Supervised Visual Representation Learning Toolbox and Benchmark

Awesome List of Masked Image Modeling (MIM) Papers for Self-supervised Visual Representation Learning

[INTERSPEECH 2024] Official pytorch code for the paper "Disentangled Representation Learning for Environment-agnostic Speaker Recognition"

PyTorch implementation of MAE https//arxiv.org/abs/2111.06377

The official repo of Qwen-Audio (通义千问-Audio) chat & pretrained large audio language model proposed by Alibaba Cloud.

ICASSP 2023: 'Speaker recognition with two-step multi-modal deep cleansing'

Code repository for FreGrad

INTERSPEECH2023: Target Active Speaker Detection with Audio-visual Cues