visual-generation 开源项目推荐（19 个）

FoundationVision /InfinityStar

[NeurIPS 2025 Oral]Infinity⭐️: Unified Spacetime AutoRegressive Modeling for Visual Generation

MiniMax-AI /VTP

[ECCV 2026] Towards Scalable Pre-training of Visual Tokenizers for Generation

Correr-Zhou /OmniShow

[ICML 2026] ByteDance's All-in-One Video Generation Model for Human-Object Interaction Video Generation

dososo /blcaptain-style-skill

中文图文卡片生成 Skill：一段内容 → 海报级图文组图（小红书 3:4 / 公众号封面）。1 引擎 + 3 视觉语言（静纸 Still Paper / 实证 Signal Proof / 图桥 Bridge Canvas），智能驱动 + 独家影像配方。

EvolvingLMMs-Lab /Evolving-Visual-Generation

[Roadmap] Visual Generation in the New Era: An Evolution from Atomic Mapping to Agentic World Modeling

HKU-MMLab /OmniX

Official implementation of "OmniX: From Unified Panoramic Generation and Perception to Graphics-Ready 3D Scenes".

thuml /Reasoning-Visual-World

Official repository for "Visual Generation Unlocks Human-Like Reasoning through Multimodal World Models", https://arxiv.org/abs/2601.19834

Gong1130 /DIRECT

[ICML 2026] Direct 3D-Aware Object Insertion via Decomposed Visual Proxies

moshehbenavraham /Napkin-AI-API

Streamlit Web App/Python CLI client for the Napkin AI API that transforms text into professional diagrams, illustrations, and visual content with 15+ built-in styles.

Manchery /awesome-visual-tokenizer

[WIP🚧] 2025 up-to-date list of resources on visual tokenizers (primarily for visual generation). Give it a star 🌟 if you find it useful.

hu-zijing /Awesome-Multimodal-RL

This repository collects research papers on reinforcement learning in generative multimodal models.

wendell0218 /FocusDiff

Official repository of the paper "FocusDiff: Advancing Fine-Grained Text-Image Alignment for Autoregressive Visual Generation through RL"

OpenRaiser /MoTiF

🔗 Bridging Modal Isolation in Interleaved Thinking: Supervising Modality Transitions via Stepwise Reinforcement

PaladinKnightMaster /ai-image-prompt-factory

Evidence-aware visual prompt compiler and experimentation lab for GPT Image 2, with historical/cultural reasoning, reference-role control, prompt decomposition, ablation testing, and visual benchmarks.

ManfredCh /happy-ai-security

Background for AI security work: visual generative models and 3D spatial construction, plus a tutorial outline. Licence: CC BY-NC-SA 4.0

DaimyoMaharajaEmpty /ai-image-to-video-generator

AI Image-To-Video Generator is a generative media application designed to transform static images into dynamic video clips using neural network algorithms. Content creators and digital artists utilize this software to animate still artwork, simulate realistic camera movements, and produce short motion graphics automatically.

dsrivastavv /DPAR

(CVPR 2026) DPAR: Dynamic Patchification for Efficient Autoregressive Visual Generation

kantkrishan0206-crypto /gen-image3.0

a powerful, large-scale, multimodal model for Text-to-Image generation.

FIicher /Regex-Visual-Generator

Générateur visuel de regex + recherche automatique // Visual regex generator + automatic search