Free-Lunch Long Video Generation via Layer-Adaptive O.O.D Correction

• URL: http://arxiv.org/abs/2603.25209v1
• Date: Thu, 26 Mar 2026 09:12:14 GMT
• Title: Free-Lunch Long Video Generation via Layer-Adaptive O.O.D Correction
• Abstract summary: We propose FreeLOC, a training-free, layer-adaptive framework for long-video inference.<n>Our approach significantly outperforms existing training-free methods, achieving state-of-the-art results in both temporal consistency and visual quality.
• Score: 12.736570247861309
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Generating long videos using pre-trained video diffusion models, which are typically trained on short clips, presents a significant challenge. Directly applying these models for long-video inference often leads to a notable degradation in visual quality. This paper identifies that this issue primarily stems from two out-of-distribution (O.O.D) problems: frame-level relative position O.O.D and context-length O.O.D. To address these challenges, we propose FreeLOC, a novel training-free, layer-adaptive framework that introduces two core techniques: Video-based Relative Position Re-encoding (VRPR) for frame-level relative position O.O.D, a multi-granularity strategy that hierarchically re-encodes temporal relative positions to align with the model's pre-trained distribution, and Tiered Sparse Attention (TSA) for context-length O.O.D, which preserves both local detail and long-range dependencies by structuring attention density across different temporal scales. Crucially, we introduce a layer-adaptive probing mechanism that identifies the sensitivity of each transformer layer to these O.O.D issues, allowing for the selective and efficient application of our methods. Extensive experiments demonstrate that our approach significantly outperforms existing training-free methods, achieving state-of-the-art results in both temporal consistency and visual quality. Code is available at https://github.com/Westlake-AGI-Lab/FreeLOC.

Related papers

• LongE2V: Long-Horizon Event-based Video Reconstruction, Prediction, and Frame Interpolation with Video Diffusion Models [14.33633321388011]
  LongE2V is a novel approach that leverages pre-trained video diffusion priors to handle event-based video reconstruction.<n>By fine-tuning agressive video models, our approach mitigates high data efficiency and superior perceptual quality.<n>Experiments on real-world benchmarks demonstrate that LongE2V outperforms state-of-the-art methods across all three tasks.
  arXiv  Detail & Related papers  (2026-07-09T17:59:52Z)
• FlowLong: Inference-time Long Video Generation via Manifold-constrained Tweedie Matching [68.01498128172214]
  We propose a novel but simple inference-time approach for long video generation that is architecture-agnostic and requires no additional training.<n>Our method generates long videos via overlapping sliding windows, where predicted clean samples from adjacent windows are blended via emphTweedie matching to enforce both textbfmanifold constraint and temporal consistency across overlap regions.<n>Our method generates videos several times longer than the native window length while outperforming both training-free and autoregressive baselines in temporal consistency and visual quality, and further extends to audio-video joint generation and text-to-3DGS without any
  arXiv  Detail & Related papers  (2026-05-20T08:55:37Z)
• Temporal In-Context Fine-Tuning for Versatile Control of Video Diffusion Models [34.131515004434846]
  We introduce Temporal In-Context Fine-Tuning (TIC-FT), an efficient approach for adapting pretrained video diffusion models to conditional generation tasks.<n>TIC-FT requires no architectural changes and achieves strong performance with as few as 10-30 training samples.<n>We validate our method across a range of tasks, including image-to-video and video-to-video generation, using large-scale base models such as CogVideoX-5B and Wan-14B.
  arXiv  Detail & Related papers  (2025-06-01T12:57:43Z)
• ViaRL: Adaptive Temporal Grounding via Visual Iterated Amplification Reinforcement Learning [68.76048244253582]
  We introduce ViaRL, the first framework to leverage rule-based reinforcement learning (RL) for optimizing frame selection in video understanding.<n>ViaRL utilizes the answer accuracy of a downstream model as a reward signal to train a frame selector through trial-and-error.<n>ViaRL consistently delivers superior temporal grounding performance and robust generalization across diverse video understanding tasks.
  arXiv  Detail & Related papers  (2025-05-21T12:29:40Z)
• Video Decomposition Prior: A Methodology to Decompose Videos into Layers [74.36790196133505]
  This paper introduces a novel video decomposition prior VDP' framework which derives inspiration from professional video editing practices.<n>VDP framework decomposes a video sequence into a set of multiple RGB layers and associated opacity levels.<n>We address tasks such as video object segmentation, dehazing, and relighting.
  arXiv  Detail & Related papers  (2024-12-06T10:35:45Z)
• SIGMA:Sinkhorn-Guided Masked Video Modeling [69.31715194419091]
  Sinkhorn-guided Masked Video Modelling ( SIGMA) is a novel video pretraining method. We distribute features of space-time tubes evenly across a limited number of learnable clusters. Experimental results on ten datasets validate the effectiveness of SIGMA in learning more performant, temporally-aware, and robust video representations.
  arXiv  Detail & Related papers  (2024-07-22T08:04:09Z)
• COVE: Unleashing the Diffusion Feature Correspondence for Consistent Video Editing [57.76170824395532]
  Video editing is an emerging task, in which most current methods adopt the pre-trained text-to-image (T2I) diffusion model to edit the source video.<n>We propose COrrespondence-guided Video Editing (COVE) to achieve high-quality and consistent video editing.<n>COVE can be seamlessly integrated into the pre-trained T2I diffusion model without the need for extra training or optimization.
  arXiv  Detail & Related papers  (2024-06-13T06:27:13Z)
• Transform-Equivariant Consistency Learning for Temporal Sentence Grounding [66.10949751429781]
  We introduce a novel Equivariant Consistency Regulation Learning framework to learn more discriminative representations for each video. Our motivation comes from that the temporal boundary of the query-guided activity should be consistently predicted. In particular, we devise a self-supervised consistency loss module to enhance the completeness and smoothness of the augmented video.
  arXiv  Detail & Related papers  (2023-05-06T19:29:28Z)
• Unsupervised Video Domain Adaptation for Action Recognition: A Disentanglement Perspective [37.45565756522847]
  We consider the generation of cross-domain videos from two sets of latent factors. TranSVAE framework is then developed to model such generation. Experiments on the UCF-HMDB, Jester, and Epic-Kitchens datasets verify the effectiveness and superiority of TranSVAE.
  arXiv  Detail & Related papers  (2022-08-15T17:59:31Z)
• Learning by Aligning Videos in Time [10.075645944474287]
  We present a self-supervised approach for learning video representations using temporal video alignment as a pretext task. We leverage a novel combination of temporal alignment loss and temporal regularization terms, which can be used as supervision signals for training an encoder network.
  arXiv  Detail & Related papers  (2021-03-31T17:55:52Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.