diffusion — AI Digest

11 мая Pixal3D: Pixel-Aligned Image-to-3D Generation Accepted at SIGGRAPH 2026 Tencent ARC Lab research
10 июл Vidu S1: Real-Time Interactive Video Generation at 42 FPS on Consumer GPUs Shengshu Technology / Tsinghua University research
15 авг Alaya-EVOKE: From Linear-Scaling Supervision to Endless World research
5 мая UniVidX: One Diffusion Backbone for RGB, Intrinsic Maps, and RGBA Video Generation research
11 мая Mean Mode Screaming: Training Pathology Fix Enables 1000-Layer Diffusion Transformers research
11 мая Flow-OPD: On-Policy Distillation Pushes GenEval +29 Points on Stable Diffusion 3.5 research
12 мая Qwen-Image-2.0: Unified Image Generation and Editing at 2K Resolution, Top-1 on AI Arena Alibaba research
14 мая Asymmetric Flow Models: SOTA 1.57 FID on ImageNet via Rank-Asymmetric Velocity Parameterization Stanford University research
24 июн Krea Releases Krea 2 Raw and Turbo Open Weights: 12B DiT Image Model Generating in 2 Seconds Krea image
4 июл Sber Open-Sources GFusion, Russia's First Diffusion Language Model Sber models-llm
6 июл MrFlow: Training-Free 10-25x Speedup for Flow-Matching Text-to-Image Models Beihang University research
12 июл Vidu S1: A Real-Time Interactive Video Generation Model ShengShu Technology research
26 авг Unlocking the Potential of Image Editing via Concept Scaling and Dense Supervision research
26 авг On-Policy Self-Distillation in Diffusion Models ByteDance Seed research
21 июн Moebius: 0.2B Lightweight Image Inpainting Framework Matches 11.9B FLUX Model Huazhong University of Science and Technology research
28 июл Self Gradient Forcing: Native Long Video Extrapolation research
6 сен LLaDA-Image: building strong image generators with fully open training recipes inclusionAI (Ant Group) research
9 сен Uno: lossless LLM speedups via diffusion-augmented autoregression MBZUAI research
10 сен Tencent open-sources AuK, a 1.5B speech foundation model for generation and editing Tencent Hunyuan audio
16 мая Orthrus: 7.8x Inference Speedup for Qwen3 via Autoregressive-Diffusion KV Sharing research
16 мая Causal Forcing++: 2-Step Distillation Enables Real-Time Interactive Video Generation Tsinghua University research
16 мая SANA-WM: Minute-Scale 720p World Modeling on a Single GPU NVIDIA research
10 июн Flow-DPPO: Principled RL Alignment for Flow Matching Image and Video Models Tencent Hunyuan research
16 июл Boogu-Image-0.1: Strong Open-Source 10B Diffusion Transformer for Text-to-Image and Editing Boogu Project research
28 июл Mage-Flow: An Efficient Native-Resolution Foundation Model for Image Generation and Editing Microsoft Research research
7 авг JoyAI-Video-Edit: Real-Time Open-Ended Video Editing with Autoregressive Diffusion JD (Jingdong) research
14 мая AnyFlow: Any-Step Video Diffusion with On-Policy Flow Map Distillation MIT / NVIDIA research
24 авг ForgeWM: Progressive Causal Training for Few-Step Action-Conditioned Video World Models research
26 авг Block3D: Efficient Text-to-3D Generation via Block-Wise Diffusion research
9 мая Cola DLM: Continuous Latent Diffusion Language Model with Competitive Scaling research
10 июн SCAIL-2: End-to-End Character Animation via In-Context Conditioning Tsinghua University research
18 июн Diffusion-Proof: Formal Theorem Proving via Diffusion Language Models research
18 июн DreamReasoner-8B: Block-Size Curriculum for Diffusion Reasoning Models research
28 июн Tencent Hunyuan Open-Sources UniRL: Unified RL Post-Training for LLMs and Diffusion Models Tencent / Hunyuan research
23 июл Mage-Flow: An Efficient Native-Resolution Foundation Model for Image Generation and Editing Microsoft research
4 авг Scaling Properties of Text Conditioning in Visual Generation ByteDance Seed research
1 сен DreamX-Creator: native joint audio-video generation at 2K resolution AMAP-ML (Alibaba) research
1 сен GenFirst: generation-first training makes end-to-end latent generative models stable ByteDance Seed research