paper-with-me

홈 › Papers

Future Video Synthesis with Object Motion Prediction

2020-04-01 · CVPR 2020 6 · Yue Wu, Rongrong Gao, Jaesik Park, Qifeng Chen

We present an approach to predict future video frames given a sequence of continuous video frames in the past. Instead of synthesizing images directly, our approach is designed to understand the complex scene dynamics by decoupling the background scene and moving objects. The appearance of the scene components in the future is predicted by non-rigid deformation of the background and affine transformation of moving objects. The anticipated appearances are combined to create a reasonable video in the future. With this procedure, our method exhibits much less tearing or distortion artifact compared to other approaches. Experimental results on the Cityscapes and KITTI datasets show that our model outperforms the state-of-the-art in terms of visual quality and accuracy.

📄 PDF Abstract BibTeX arXiv:2004.00542

Code (1)

YueWuHKUST/FutureVideoSynthesis 공식 구현 pytorch

Tasks

motion predictionObjectPredict Future Video FramesPredictionVideo Prediction

Similar Papers 제목 키워드 기반

3D Motion Decomposition for RGBD Future Dynamic Scene Synthesis

2019-06-01 · CVPR 2019 6 · Xiaojuan Qi, Zhengzhe Liu, Qifeng Chen, Jiaya Jia

A future video is the 2D projection of a 3D scene with predicted camera and object motion. Accurate future video prediction inherently requires understanding of 3D motion and geometry of a scene. In this paper, we propos…

Video Prediction

WALDO: Future Video Synthesis using Object Layer Decomposition and Parametric Flow Prediction

2022-11-25 · ICCV 2023 1 · Guillaume Le Moing, Jean Ponce, Cordelia Schmid

This paper presents WALDO (WArping Layer-Decomposed Objects), a novel approach to the prediction of future video frames from past ones. Individual images are decomposed into multiple layers combining object masks and a s…

SSIM

Forecasting Future Videos from Novel Views via Disentangled 3D Scene Representation

2024-07-31 · Sudhir Yarram, Junsong Yuan

Video extrapolation in space and time (VEST) enables viewers to forecast a 3D scene into the future and view it from novel viewpoints. Recent methods propose to learn an entangled representation, aiming to model layered …

Motion ForecastingNovel View Synthesis

GaussianPrediction: Dynamic 3D Gaussian Prediction for Motion Extrapolation and Free View Synthesis

2024-05-30 · Boming Zhao, Yuan Li, Ziyu Sun, Lin Zeng 외

Forecasting future scenarios in dynamic environments is essential for intelligent decision-making and navigation, a challenge yet to be fully realized in computer vision and robotics. Traditional approaches like video pr…

Decision MakingNovel View SynthesisVideo Prediction

State-space Decomposition Model for Video Prediction Considering Long-term Motion Trend

2024-04-17 · Fei Cui, Jiaojiao Fang, Xiaojiang Wu, Zelong Lai 외

Stochastic video prediction enables the consideration of uncertainty in future motion, thereby providing a better reflection of the dynamic nature of the environment. Stochastic video prediction methods based on image au…

motion predictionPredictionState Space ModelsVideo Prediction