computer-vision-4

Perpetual View Generation

3260 papers • 126 benchmarks • 313 datasets

Perpetual View Generation is the task of generating long-range novel views by flying into a given image.

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in perpetual-view-generation-4

Trend

Dataset

Best Model

Actions

LHQ

Libraries

i

Use these libraries to find perpetual-view-generation-4 models and implementations

Datasets

LHQ

Subtasks

No subtasks available.

Most implemented papers

Infinite Nature: Perpetual View Generation of Natural Scenes from a Single Image

Noah Snavely, Angjoo Kanazawa, Richard Tucker, Andrew Liu, V. Jampani, A. Makadia•Wed Dec 16 2020

This work introduces the problem of perpetual view generation— long-range generation of novel views corresponding to an arbitrarily long camera trajectory given a single image, and takes a hybrid approach that integrates both geometry and image synthesis in an iterative ‘render, refine and repeat’ framework.

240

Content

0

Paper Graph

InfiniteNature-Zero: Learning Perpetual View Generation of Natural Scenes from Single Images

Noah Snavely, Angjoo Kanazawa, Zhengqi Li, Qianqian Wang•Thu Jul 21 2022

A novel self-supervised view generation training paradigm is proposed, where the model is sample and rendering virtual camera trajectories, including cyclic ones, allowing the model to learn stable view generation from a collection of single views.

82 0

Paper Graph

Adding a benchmark result helps the community track progress.