Public GitHub footprint of KlingAI Research

Bring portraits to life!

[ICCV'25 Best Paper Finalist] ReCamMaster: Camera-Controlled Generative Rendering from A Single Video

[ICLR'25] SynCamMaster: Synchronizing Multi-Camera Video Generation from Diverse Viewpoints

[ICLR 2026] UniVideo: Unified Understanding, Generation, and Editing for Videos

[ICCV 2025] GameFactory: Creating New Games with Generative Interactive Videos

[NeurIPS 2025] Improving Video Generation with Human Feedback

[ICLR'25] 3DTrajMaster: Mastering 3D Trajectory for Multi-Entity Motion in Video Generation

Official implementation of the paper "Koala-36M: A Large-scale Video Dataset Improving Consistency between Fine-grained Conditions and Video Content".

I2V-Adapter: A General Image-to-Video Adapter for Diffusion Models

Official Implementation of "MemFlow: Flowing Adaptive Memory for Consistent and Efficient Long Video Narratives"

Try X-Dub to sync any character in a video with any audio you like | Official repository for "From Inpainting to Editing: Unlocking Robust Mask-Free Visual Dubbing via Generative Bootstrapping"

[Arxiv 2025] Official PyTorch implementation of DiffMoE, TC-DiT, EC-DiT and Dense DiT

[CVPR'25] StyleMaster: Stylize Your Video with Artistic Generation and Translation

No description provided for this repository.

CVPR 2026 | Official Implementation of "MultiShotMaster: A Controllable Multi-Shot Video Generation Framework"

[SIGGRAPH Asia'25] Enabling Reference-based Camera Control via Context without Explicit 3D Estimation

[Arxiv 2025] Official PyTorch Implementation of "SVG-T2I: Scaling up Text-to-Image Latent Diffusion Model Without Variational Autoencoder".

ShotStream: Streaming Multi-Shot Video Generation for Interactive Storytelling

[CVPR 2026] Video-as-Answer: Predict and Generate Next Video Event with Joint-GRPO

Official implementation of "HumanAesExpert: Advancing a Multi-Modality Foundation Model for Human Image Aesthetic Assessment"

The official implementation of StereoPilot

[ICLR’26] Learning Video Generation for Robotic Manipulation with Collaborative Trajectory Control

Unified Multi-modal IAA Baseline and Benchmark

Official Code of "VideoCanvas: Unified Video Completion from Arbitrary Spatiotemporal Patches via In-Context Conditioning"

[ICML 2025 Spotlight] MODA: MOdular Duplex Attention for Multimodal Perception, Cognition, and Emotion Understanding

Official Pytorch implementation of AvatarForcing: One-Step Streaming Talking Avatars via Local-Future Sliding-Window Denoising

Official implementation of paper "VMoBA: Mixture-of-Block Attention for Video Diffusion Models"

Official implementation of "SPF-Portrait: Towards Pure Portrait Customization with Semantic Pollution-Free Fine-tuning"

Official repository of PhysMaster: Mastering Physical Representation for Video Generation via Reinforcement Learning

[ICLR'26] Easier Painting Than Thinking: Can Text-to-Image Models Set the Stage, but Not Direct the Play?

No description provided for this repository.

Metric implementation and raw data of "Diffusing in the Right Space: A Systematic Study of Latent Diffusability"

DecMem: Towards Minute-Long Consistent World Generation with Decoupled Memory

[NeurIPS'25] VidEmo: Affective-Tree Reasoning for Emotion-Centric Video Foundation Models

[ACL'26 Oral] Official implementation of "SegTune: Structured and Fine-Grained Control for Song Generation".

VQRAE: Representation Quantization Autoencoders for Multimodal Understanding, Generation and Reconstruction

Official implementation of NeurIPS'25 paper "VFRTok: Variable Frame Rates Video Tokenizer with Duration-Proportional Information Assumption"

[ICCV 2025] Official Implementation of the Paper "Imbalance in Balance: Online Concept Balancing in Generation Models".

Decoupled Video Instance Segmentation Framework, improved version of dvis

TexEditor: Structure-Preserving Text-Driven Texture Editing

Decoupled Video Instance Segmentation Framework

[ICLR 2026] Scalingcache: extreme acceleration of dits through difference scaling and dynamic interval caching

This is the program for supporting KlingAI Express in WAIC 2025.

Scripts for processing and evaluating SocioEmoDialog datasets. It includes the core processing scripts, evaluation metrics, and additional documentation.

No description provided for this repository.

RewardHarness: Self-Evolving Agentic Post-Training https://rewardharness.com/

No description provided for this repository.

No description provided for this repository.

No description provided for this repository.

Making Image Editing Easier via Adaptive Task Reformulation with Agentic Executions

[ICLR'26] Mitigating the Noise Shift for Denoising Generative Models via Noise Awareness Guidance