Papers › Latent Video Diffusion Models for High-Fidelity Long Video Generation

Latent Video Diffusion Models for High-Fidelity Long Video Generation

23 Nov 2022arXiv:2211.13221archive 2025-07-28

Yingqing He, Tianyu Yang, Yong Zhang, Ying Shan, Qifeng Chen

AI-generated content has attracted lots of attention recently, but photo-realistic video synthesis is still challenging. Although many attempts using GANs and autoregressive models have been made in this area, the visual quality and length of generated videos are far from satisfactory. Diffusion models have shown remarkable results recently but require significant computational resources. To address this, we introduce lightweight video diffusion models by leveraging a low-dimensional 3D latent space, significantly outperforming previous pixel-space video diffusion models under a limited computational budget. In addition, we propose hierarchical diffusion in the latent space such that longer videos with more than one thousand frames can be produced. To further overcome the performance degradation issue for long video generation, we propose conditional latent perturbation and unconditional guidance that effectively mitigate the accumulated errors during the extension of video length. Extensive experiments on small domain datasets of different categories suggest that our framework generates more realistic and longer videos than previous strong baselines. We additionally provide an extension to large-scale text-to-video generation to demonstrate the superiority of our work. Our code and models will be made publicly available.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

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

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 14 code samples harvested from 1 repository linked to this paper: 13 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 1 honoured, 1 violated, 4 with no contract checked; 7 where Syntology's instrument failed) · 1 unverified. Where Syntology's instrument failed: 6 where our contract draft was wrong; 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 14 samples from 1 repository, 13 ran (6 with no instrument failure; 7 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written between 2026-09-22 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 3 of the 13 samples that ran have no recorded date.

yingqinghe/lvdm officialmentioned in papermentioned on GitHubpytorchMITSyntology: 14 samples, 13 ran (7 instrument-failed) · 1 unverified report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

14 samples harvested: 13 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 1 honoured, 1 violated, 4 with no contract checked; 7 where Syntology's instrument failed) · 1 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · honoured contract
1ran · violated contract
6ran · our draft was wrong
1ran; the contract check could not run
4ran
1unverified

Licence: 5 of the 14 samples are pointer only, meaning Syntology does not serve that copy's text. For 5 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from yingqinghe/lvdm. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

yingqinghe/lvdm official repository 14 samples, 13 ran (7 instrument-failed) · 1 unverified
Normalize in 154 papers lvdm/models/modules/aemodules.py ran · our draft was wrong MIT (permissive) · 9fcdaa6e423e8aa7 · report fixture: signature in_channels:int/0/int, num_groups:int/0/int · returned GroupNorm · level recorded 2026-09-22
adopt_weight in 12 papers lvdm/models/losses/vqperceptual.py ran · honoured contract fingerprinted MIT (permissive) here; the stored copy is pointer only · 7439a9d4c6be04b0 · report fixture: signature weight:float/0/float, global_step:int/0/int, threshold:int/0/int, value:float/0/float · returned float · level recorded 2026-09-22
disabled_train in 172 papers lvdm/models/ddpm3d.py ran · violated contract MIT (permissive) · 4cb732f513d69dfd · report fixture: signature self:callable/0/?, mode:bool/0/bool · returned SimpleModel · level recorded, date not recorded
hinge_d_loss in 19 papers lvdm/models/modules/aemodules3d.py ran · our draft was wrong fingerprinted MIT (permissive) here; the stored copy is pointer only · 22a83df891d750c6 · report fixture: signature logits_real:Tensor/2/float, logits_fake:Tensor/2/float · returned Tensor · level recorded 2026-09-22
hinge_d_loss_with_exemplar_weights lvdm/models/losses/vqperceptual.py ran MIT (permissive) · af24442fe485c871 · report fixture: signature logits_real:Tensor/4/float, logits_fake:Tensor/4/float, weights:Tensor/1/float · returned Tensor · level recorded 2026-09-24
make_attn in 22 papers lvdm/models/modules/aemodules.py ran MIT (permissive) · ff5ec4cc14b9e9b2 · report fixture: signature in_channels:int/0/int, attn_type:str/0/str · returned AttnBlock · level recorded 2026-09-25
measure_perplexity lvdm/models/losses/vqperceptual.py ran MIT (permissive) · 105b7d55d7897384 · report fixture: signature predicted_indices:Tensor/1/int, n_embed:int/0/int · returned tuple · level recorded 2026-09-25
nonlinearity in 140 papers
census: 4 implementations from 11 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
lvdm/models/modules/aemodules.py ran · our draft was wrong fingerprinted MIT (permissive) · 3137073275f8c21a · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded, date not recorded
normal_kl in 188 papers
census: 14 implementations from 24 papers, 2 distinct outputs on one shared input
lvdm/models/modules/distributions.py ran; the contract check could not run (RAISES) fingerprinted MIT (permissive) here; the stored copy is pointer only · 17eb0a31c91b671c · report fixture: signature mean1:Tensor/2/float, logvar1:Tensor/2/float, mean2:Tensor/2/float, logvar2:Tensor/2/float · returned Tensor · level recorded 2026-09-24
silu in 15 papers
census: 5 implementations from 7 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
lvdm/models/modules/aemodules3d.py ran · our draft was wrong fingerprinted MIT (permissive) here; the stored copy is pointer only · a8c56a46a8cde9c8 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-22
split_video_to_clips lvdm/models/ddpm3d.py ran MIT (permissive) · 9a19a68d5b2f9a37 · report fixture: signature video:Tensor/5/float, clip_length:int/0/int, drop_left:bool/0/bool · returned Tensor · level recorded 2026-09-25
uniform_on_device in 45 papers lvdm/models/ddpm3d.py ran · our draft was wrong MIT (permissive) · d48d8354986e3b0e · report fixture: signature r1:float/0/float, r2:float/0/float, shape:tuple/1/int, device:device/0/? · returned Tensor · level recorded, date not recorded
vanilla_d_loss in 16 papers lvdm/models/modules/aemodules3d.py ran · our draft was wrong fingerprinted MIT (permissive) here; the stored copy is pointer only · 263c54290c553ddd · report fixture: signature logits_real:Tensor/1/float, logits_fake:Tensor/1/float · returned Tensor · level recorded 2026-09-22
make_spatialtemporal_transformer lvdm/models/modules/openaimodel3d.py unverified MIT (permissive) · b762ab98954e114e · report

Tasks

DenoisingImage GenerationText-to-Video GenerationVideo GenerationVocal Bursts Intensity Prediction

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Video Generation Sky Time-lapse LVDM (256x256) FVD 16 95.2 #4 of 9 Archive leaderboard report
Video Generation Sky Time-lapse LVDM (256x256) KVD16 3.9 #4 of 9 Archive leaderboard report
Video Generation Sky Time-lapse Long-video GAN (128x128) FVD 16 107.5 #5 of 9 Archive leaderboard report
Video Generation Sky Time-lapse DIGAN (128x128) FVD 16 114.6 #6 of 9 Archive leaderboard report
Video Generation Sky Time-lapse DIGAN (128x128) KVD16 6.8 #6 of 9 Archive leaderboard report
Video Generation Sky Time-lapse Long-video GAN (256x256) FVD 16 116.5 #7 of 9 Archive leaderboard report
Video Generation Sky Time-lapse TATS (128x128) FVD 16 132.6 #8 of 9 Archive leaderboard report
Video Generation Sky Time-lapse TATS (128x128) KVD16 5.7 #8 of 9 Archive leaderboard report
Video Generation Sky Time-lapse MoCoGAN-HD (128x128) FVD 16 183.6 #9 of 9 Archive leaderboard report
Video Generation Sky Time-lapse MoCoGAN-HD (128x128) KVD16 13.9 #9 of 9 Archive leaderboard report
Video Generation Taichi TATS (128x128) FVD16 94.6 #2 of 6 Archive leaderboard report
Video Generation Taichi TATS (128x128) KVD16 9.8 #2 of 6 Archive leaderboard report
Video Generation Taichi LVDM (256x256) FVD16 99 #3 of 6 Archive leaderboard report
Video Generation Taichi LVDM (256x256) KVD16 15.3 #3 of 6 Archive leaderboard report
Video Generation Taichi DIGAN (128x128) FVD16 128.1 #4 of 6 Archive leaderboard report
Video Generation Taichi DIGAN (128x128) KVD16 20.6 #4 of 6 Archive leaderboard report
Video Generation Taichi MoCoGAN-HD (128x128) FVD16 144.7 #5 of 6 Archive leaderboard report
Video Generation Taichi MoCoGAN-HD (128x128) KVD16 25.4 #5 of 6 Archive leaderboard report
Video Generation Taichi DIGAN (256x256) FVD16 156.7 #6 of 6 Archive leaderboard report
Video Generation UCF-101 LVDM (256x256, unconditional) FVD16 372 #32 of 48 Archive leaderboard report
Video Generation UCF-101 LVDM (256x256, unconditional) KVD16 27 #32 of 48 Archive leaderboard report
Video Generation UCF-101 LVDM (256x256, unconditional) FVD16 552 #39 of 48 Archive leaderboard report
Video Generation UCF-101 LVDM (256x256, unconditional) KVD16 42 #39 of 48 Archive leaderboard report
Video Generation UCF-101 TGAN-v2 (128x128) FVD16 1209 #45 of 48 Archive leaderboard report
Video Generation UCF-101 VDM FVD16 1396 #46 of 48 Archive leaderboard report
Video Generation UCF-101 VDM KVD16 116 #46 of 48 Archive leaderboard report
Video Generation UCF-101 MCVD FVD16 2460 #47 of 48 Archive leaderboard report
Video Generation UCF-101 MCVD KVD16 148 #47 of 48 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

Diffusion

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