# DEVIS-GRPO: Unleashing GRPO on Dynamic Extreme View Synthesis

arXiv: https://arxiv.org/abs/2605.16937 · HTML: https://mlsanity.com/paper/2605.16937/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 37 of its 47 references that are in the map.
- First posted on arXiv 16 May 2026.

## Builds on (most cited first, 30 of 37)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [3D Gaussian Splatting for Real-Time Radiance Field Rendering](https://mlsanity.com/paper/2308.04079/) (2023)
- [Instant Neural Graphics Primitives with a Multiresolution Hash Encoding](https://mlsanity.com/paper/2201.05989/) (2022)
- [Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets](https://mlsanity.com/paper/2311.15127/) (2023)
- [CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer](https://mlsanity.com/paper/2408.06072/) (2024)
- [VBench: Comprehensive Benchmark Suite for Video Generative Models](https://mlsanity.com/paper/2311.17982/) (2023)
- [DUSt3R: Geometric 3D Vision Made Easy](https://mlsanity.com/paper/2312.14132/) (2023)
- [AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models without Specific Tuning](https://mlsanity.com/paper/2307.04725/) (2023)
- [4D Gaussian Splatting for Real-Time Dynamic Scene Rendering](https://mlsanity.com/paper/2310.08528/) (2023)
- [Grounding Image Matching in 3D with MASt3R](https://mlsanity.com/paper/2406.09756/) (2024)
- [Training Diffusion Models with Reinforcement Learning](https://mlsanity.com/paper/2305.13301/) (2023)
- [WorldSimBench: Towards Video Generation Models as World Simulators](https://mlsanity.com/paper/2410.18072/) (2024)
- [Flow-GRPO: Training Flow Matching Models via Online RL](https://mlsanity.com/paper/2505.05470/) (2025)
- [GAIA-1: A Generative World Model for Autonomous Driving](https://mlsanity.com/paper/2309.17080/) (2023)
- [MotionCtrl: A Unified and Flexible Motion Controller for Video Generation](https://mlsanity.com/paper/2312.03641/) (2023)
- [DL3DV-10K: A Large-Scale Scene Dataset for Deep Learning-based 3D Vision](https://mlsanity.com/paper/2312.16256/) (2023)
- [VideoCrafter1: Open Diffusion Models for High-Quality Video Generation](https://mlsanity.com/paper/2310.19512/) (2023)
- [Sora: A Review on Background, Technology, Limitations, and Opportunities of Large Vision Models](https://mlsanity.com/paper/2402.17177/) (2024)
- [Real-time Photorealistic Dynamic Scene Representation and Rendering with 4D Gaussian Splatting](https://mlsanity.com/paper/2310.10642/) (2023)
- [CameraCtrl: Enabling Camera Control for Text-to-Video Generation](https://mlsanity.com/paper/2404.02101/) (2024)
- [Lumiere: A Space-Time Diffusion Model for Video Generation](https://mlsanity.com/paper/2401.12945/) (2024)
- [DanceGRPO: Unleashing GRPO on Visual Generation](https://mlsanity.com/paper/2505.07818/) (2025)
- [ReCamMaster: Camera-Controlled Generative Rendering from A Single Video](https://mlsanity.com/paper/2503.11647/) (2025)
- [DepthSplat: Connecting Gaussian Splatting and Depth](https://mlsanity.com/paper/2410.13862/) (2024)
- [Monocular Dynamic View Synthesis: A Reality Check](https://mlsanity.com/paper/2210.13445/) (2022)
- [4D-Rotor Gaussian Splatting: Towards Efficient Novel View Synthesis for Dynamic Scenes](https://mlsanity.com/paper/2402.03307/) (2024)
- [MV-DUSt3R+: Single-Stage Scene Reconstruction from Sparse Views In 2 Seconds](https://mlsanity.com/paper/2412.06974/) (2024)

## Abstract (arXiv)

Trajectory-controlled video generation has become essential for controllable video generation. While current methods perform well under small-view camera motions, they degrade significantly with large-view motions. Existing solutions for extreme-view synthesis typically require dedicated video pairs, demanding substantial annotation effort. To address these limitations, we propose Dynamic Extreme VIew Synthesis-GRPO (DEVIS-GRPO), a GRPO-based framework for trajectory-controlled video generation, the first online policy gradient method for extreme view video generation. Central to our approach is a novel sampling strategy: Accumulative Dynamic Extreme VIew Synthesis (ADEVIS), which achieves large-view camera motions by progressively accumulating small-view increments. This method delivers two key advantages: 1) enhanced training efficiency, as it eliminates the need to warm-start the policy model by collecting expensive paired large-view videos, and 2) increased sampling diversity, achieved by flexibly varying trajectory configurations. Finally, we designed a multi-level consistency-quality reward function to select high-quality samples for model optimization. Experiments on the Kubric-4D, iPhone, and DL3DV datasets demonstrate our method's superiority. On Kubric-4D, we achieve relative improvements of 21.57% in PSNR and 7.31% in SSIM over the second-best method in non-occlusion areas. On iPhone, LPIPS is reduced by 18.56%.