Jun 2026· arXiv.org· Vol abs/2606.29198· 0 citations· 52 references
Computer Science
TL;DR
This work presents Dynamic Trajectory Initialization (DTI) paradigm for GFVSR, which reformulates GFVSR as an input-driven directional restoration and demonstrates the perception-distortion trade-off and that the LPIPS is the most convincing metric in this case.
Abstract
As the most perceptually powerful Face Video Super-Resolution (FVSR) method, existing works in Generative FVSR (GFVSR) mainly exploit the generative prior of pretrained diffusion models. However, viewed as full generation, they suffer from fixed sampling and expensive inference costs if without large-scale auxiliary training. Furthermore, an excessive pursuit of generic perceptual metrics often results in low fidelity. To address these issues, we present Dynamic Trajectory Initialization (DTI) paradigm for GFVSR, which reformulates GFVSR as an input-driven directional restoration. With a novel enhancement-and-injection conditioning mechanism for pretrained DiT backbone, fidelity of our model has been significantly improved without compromising perceptual quality. To dynamically set the starting sampling point, we propose a Discriminative Guide (DG) trained via objective Signal-to-Noise Ratio (SNR) alignment. With only minor model adaptation and fine-tuning, our method achieves a SOTA overall performance across diverse metrics and benchmarks. An analysis of relationship between actual comprehensive quality and common metrics is also conducted, which demonstrates the perception-distortion trade-off and that the LPIPS is the most convincing metric in our case.
Despite the success of diffusion models in Video Frame Interpolation (VFI), existing methods still suffer from two critical limitations. First, latent diffusion inevitably loses fine-grained details when reconstructing images from latent representations back to the pixel space. Second, multi-step sampling incurs prohibitive memory consumption and inference latency. To address these issues, we propose SPEED, a one-step pixel diffusion framework for high-quality VFI. Specifically, SPEED employs a progressive multi-stage architecture with dynamic patch scaling to effectively learn multi-scale motion, structural, and appearance representations. Furthermore, we propose a novel Noise-Update-Only Attention mechanism to prevent semantic degradation of the clean condition frames while reducing the computational overhead by nearly 50%. Besides, we introduce a Drift-aware Timestep Sampling strategy coupled with a tailored training objective to directly predict images in the pixel space, enabling one-step inference without compromising the quality of the generated frames. Extensive experiments show that SPEED achieves state-of-the-art performance. On SNU-FILM, SPEED reduces LPIPS by 8.8% while delivering 63.3% faster inference and 10.6% lower memory usage. On challenging 4K benchmarks, it further surpasses prior methods by up to 51.5% in LPIPS.
Zihao Zhang, Haoyu Zhao, Siqian Yang et al.· 0 citations
Residual Flow Matching for Image Super-Resolution (RFMSR) is proposed, a vision-only framework that centers the source distribution at the LQ latent, reducing transport distance and preserving structural priors throughout the flow trajectory.
Shuwei Huang, Tianyao Luo, Jicheng Liu et al.· 1 citation
This work proposes MeanSR, a one-step perceptual SR method that learns an LR-conditioned average velocity field to directly capture the finite-time transition from degraded or noisy inputs to plausible HR outputs and introduces a Stage-Aware Temporal Sampling strategy to improve trajectory learning.
Axi Niu, Jiawei Kou, Kang Zhang et al.· 0 citations
MotionCraft is presented, a controllable VSR framework that formulates restoration as motion-aware latent state prediction inspired by world models and integrates adaptive sparse attention with an explicit user-accessible control interface to deliver temporally consistent, high-quality reconstructions under streaming constraints.
Rong Fu, Chunlei Meng, Yangcheng Zeng et al.· 0 citations
FADRA, a frequency-aware diffusion framework with iterative residual adaptation specifically tailored for robust VFR, is proposed and a Frequency-Aware Loss is introduced that provides explicit supervision across multiple spectral bands, emphasizing visually sensitive frequency components that are crucial for perceptual quality and prone to temporal jittering.
Jin Jiang, Jia Wang, Panwen Hu et al.· 0 citations
A diffusion-based framework with a stage-wise multi-condition guidance mechanism that enhances both structural and textural fidelity and compares with recent image-to-image translation and diffusion-based baselines to observe competitive performance in both visual coherence and identity preservation.
Yue Que, Xuegui Cheng, Shuqian Shi et al.· Neural Networks· 0 citations