Papers › DiffDreamer: Towards Consistent Unsupervised Single-view Scene Extrapolation with...

DiffDreamer: Towards Consistent Unsupervised Single-view Scene Extrapolation with Conditional Diffusion Models

22 Nov 2022ICCV 2023 1arXiv:2211.12131archive 2025-07-28

Shengqu Cai, Eric Ryan Chan, Songyou Peng, Mohamad Shahbazi, Anton Obukhov, Luc van Gool, Gordon Wetzstein

Scene extrapolation -- the idea of generating novel views by flying into a given image -- is a promising, yet challenging task. For each predicted frame, a joint inpainting and 3D refinement problem has to be solved, which is ill posed and includes a high level of ambiguity. Moreover, training data for long-range scenes is difficult to obtain and usually lacks sufficient views to infer accurate camera poses. We introduce DiffDreamer, an unsupervised framework capable of synthesizing novel views depicting a long camera trajectory while training solely on internet-collected images of nature scenes. Utilizing the stochastic nature of the guided denoising steps, we train the diffusion models to refine projected RGBD images but condition the denoising steps on multiple past and future frames for inference. We demonstrate that image-conditioned diffusion models can effectively perform long-range scene extrapolation while preserving consistency significantly better than prior GAN-based methods. DiffDreamer is a powerful and efficient solution for scene extrapolation, producing impressive results despite limited supervision. Project page: https://primecai.github.io/diffdreamer.

PaperPDFConference PDF

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

No code repository is listed for this paper in the archive or in Syntology's graph.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

DenoisingPerpetual View Generation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Perpetual View Generation LHQ DiffDreamer FID (first 20 steps) 34.49 #1 of 2 Archive leaderboard report
Perpetual View Generation LHQ DiffDreamer FID (full 100 steps) 51 #1 of 2 Archive leaderboard report
Perpetual View Generation LHQ DiffDreamer IS (first 20 steps) 2.82 #1 of 2 Archive leaderboard report
Perpetual View Generation LHQ DiffDreamer IS (full 100 steps) 2.99 #1 of 2 Archive leaderboard report
Perpetual View Generation LHQ DiffDreamer KID (first 20 steps) 0.08 #1 of 2 Archive leaderboard report
Perpetual View Generation LHQ DiffDreamer KID (full 100 steps) 0.28 #1 of 2 Archive leaderboard report
Perpetual View Generation LHQ InfNat-Zero FID (first 20 steps) 39.45 #2 of 2 Archive leaderboard report
Perpetual View Generation LHQ InfNat-Zero FID (full 100 steps) 26.24 #2 of 2 Archive leaderboard report
Perpetual View Generation LHQ InfNat-Zero IS (first 20 steps) 2.8 #2 of 2 Archive leaderboard report
Perpetual View Generation LHQ InfNat-Zero IS (full 100 steps) 2.72 #2 of 2 Archive leaderboard report
Perpetual View Generation LHQ InfNat-Zero KID (first 20 steps) 0.12 #2 of 2 Archive leaderboard report
Perpetual View Generation LHQ InfNat-Zero KID (full 100 steps) 0.12 #2 of 2 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

DiffusionInpainting

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections