NTIRE 2025 Challenge on Short-form UGC Video Quality Assessment and Enhancement: Methods and Results

2025-04-17

AI Summary

Practical deployment of image quality assessment (IQA) and enhancement for short-form user-generated content (S-UGC) on platforms like Kwai and TikTok faces bottlenecks due to computational overhead and lack of representative benchmarks. Method: (1) We introduce KwaiSR—the first large-scale, S-UGC-oriented super-resolution dataset, comprising both synthetic and real-world short-video images; (2) we propose a novel, ensemble-free, low-parameter visual quality assessment (VQA) paradigm that eliminates reliance on redundant backbone networks typical in conventional IQA/VQA; (3) we integrate a lightweight neural network with a diffusion model to achieve end-to-end, quality-aware image super-resolution. Contribution/Results: Our framework advances the practical deployment of S-UGC quality assessment and enhancement, significantly improving perceptual quality and user experience. The code and KwaiSR dataset are publicly released, catalyzing a challenge with 266 participating teams; 18 submitted solutions empirically validate the effectiveness and generalizability of our approach.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Knowledge Representation and Reasoning: Qualitative Reasoning
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

This paper presents a review for the NTIRE 2025 Challenge on Short-form UGC Video Quality Assessment and Enhancement. The challenge comprises two tracks: (i) Efficient Video Quality Assessment (KVQ), and (ii) Diffusion-based Image Super-Resolution (KwaiSR). Track 1 aims to advance the development of lightweight and efficient video quality assessment (VQA) models, with an emphasis on eliminating reliance on model ensembles, redundant weights, and other computationally expensive components in the previous IQA/VQA competitions. Track 2 introduces a new short-form UGC dataset tailored for single image super-resolution, i.e., the KwaiSR dataset. It consists of 1,800 synthetically generated S-UGC image pairs and 1,900 real-world S-UGC images, which are split into training, validation, and test sets using a ratio of 8:1:1. The primary objective of the challenge is to drive research that benefits the user experience of short-form UGC platforms such as Kwai and TikTok. This challenge attracted 266 participants and received 18 valid final submissions with corresponding fact sheets, significantly contributing to the progress of short-form UGC VQA and image superresolution. The project is publicly available at https://github.com/lixinustc/KVQE- ChallengeCVPR-NTIRE2025.

Problem

• Develop lightweight video quality assessment models
• Enhance short-form UGC videos via super-resolution
• Improve user experience on platforms like TikTok

Innovation

• Lightweight efficient video quality assessment models
• Diffusion-based super-resolution for UGC images
• New dataset for short-form UGC enhancement