cs.CVNov 13, 2026

Towards Seconds-level Video Representation: Quadratically Gated Feature Grid with Adaptive, Prediction-free Sampling

Authors: Tie-Yun Li, Hao-Ran Wang, Chao Wang, Zhiyong Yin, Yiming Xu, Zhenwei Cai, Li Li, Rongzhi Lin, 

Organizations: School of Mathematics, Harbin Institute of Technology, Harbin 150001, China. · Frontier of Artificial Networks (FAN) Lab, Department of Data Science, City University of Hong Kong, Hong Kong, China SAR. · China Construction Third Engineering Bureau (Shenzhen) Co., Ltd., Shenzhen, Guangdong 518110, China. · MoE Key Laboratory of Brain-Inspired Intelligent Perception and Cognition, University of Science and Technology of China, Hefei 230027, China.

Abstract

Implicit neural video representations are notoriously slow to build, which does not cross the basic practical threshold for most industrial applications. Here, we present Q-GAS, a coordinate-queryable video representation that couples multi-resolution explicit 3D feature grids with a compact SwiGLU-style quadratic-gating (QG) decoder and an error-aware, prediction-free sampler. The explicit grids avoid hash-collision coefficient sharing at the cost of modestly higher storage yet greatly boost the expressive capacity; the QG decoder introduces low-rank local second-order interactions to improve the nonlinear feature response. During fitting, a global error map periodically refreshes a high-error pool, and each batch mixes pool samples with uniform samples, which is free from the auxiliary model to predict difficult samples, thereby further escalating the efficiency. On UVG Beauty/Bosphorus/Jockey, Q-GAS fits and quantizes under 50 s on one RTX PRO 6000, decodes at over 370 FPS with mixed precision and about 146 FPS in FP32. In our PyTorch implementation, Q-GAS trains and quantizes 5.5-7.2 times faster than the evaluated Instant-NGP baseline and 19-28 times faster than RNeRV. Under a 90 s budget, it improves PSNR/MS-SSIM/LPIPS over the evaluated NeRV-family pipelines at comparable BPP over the measured overlap. We have shared our code at https://github.com/MillenRosen/Q-GAS for readers' free use.

Figures & tables

Explore similar work

CardsList
  1. HyperGS: Fast and Generalizable Gaussian Video Representation

    Jul 13, 2026Fatimah Zohra, Chen Zhao, Shuming Liu +23D Gaussian SplattingSpatiotemporal Transformers

  2. Video Compression with Graph-inspired Neural Representation

    Sep 2, 2026Changqi Wang, Ge Gao, Fan Zhang +4Neural Video CompressionImplicit Neural Representations

  3. STGV: Spatio-Temporal Hash Encoding for Gaussian-based Video Representation

    Apr 13, 2026Jierun Lin, Jiacong Chen, Qingyu Mao +4Video Representation Learning3D Gaussian Splatting