Nov 2026· Proceedings of the 34th ACM International Conference on Multimedia· pp. 17686-17692· 0 citations· 39 references
TL;DR
This work model machine oriented quality as the task utility retained after visual distortion degrades useful evidence, and combines signal degradation, structural visibility, semantic consistency, and action critical cues, with score calibration for correlation based evaluation.
Abstract
As visual data are increasingly consumed by recognition models, vision language systems, and robotic agents, perceptual quality alone cannot fully reflect downstream utility. Mild visual distortions may remove boundary, color, semantic, or affordance cues important to machine perception, while stronger perceptual degradation may still preserve enough information for recognition or action. We model machine oriented quality as the task utility retained after visual distortion degrades useful evidence. The model combines signal degradation, structural visibility, semantic consistency, and action critical cues, with score calibration for correlation based evaluation. The same formulation is used for absolute machine vision utility and embodied action utility. Reference relative quality is derived from paired absolute utility estimates of distorted and clean images. We implement the method using frozen visual representations, lightweight regressors, reliability score aggregation, and calibration, without fine tuning large vision language backbones. On the ACM Multimedia 2026 M-VQA Challenge, covering VLM performance prediction, relative degradation assessment, and embodied action utility prediction, the final system achieves official scores of 0.7390, 0.9065, and 0.4739, respectively, and ranks first on Track~2.
With the development of Embodied AI, machines have replaced humans as the main consumers of visual media, yet existing Image Quality Assessment (IQA) metrics remain focused on humans and overlook machine preference (a domain where quality depends on task utility rather than perceptual fidelity alone). To address this g...
Chunyi Li, Rui Qing, Jian-Bo Zhang et al.· Proceedings of the 34th ACM...· 0 citations
In Embodied AI, image quality assessment (IQA) should move beyond human visual preference and estimate whether degraded visual inputs remain useful for real-world task execution. Real-world degradations may affect task-relevant visual evidence, multimodal reasoning, and action-related decision making, making convention...
Jun-Feng Yang, Jia-Liang Wang, Xiao-Long Li et al.· Proceedings of the 34th ACM...· 0 citations
World Action Models (WAMs) inherit powerful priors from pretrained video generation models, but their variational autoencoder (VAE) representations are optimized for visual reconstruction rather than manipulation. External visual representations can improve manipulation, but existing methods couple representation choic...
Kai Liu, Jiangran Lyu, Ting-Yu Deng et al.· 2 citations
Reliable blind image-quality assessment is essential for visual monitoring of power-field operations, where images captured by fixed cameras, unmanned aerial vehicles, and handheld terminals are frequently affected by blur, noise, uneven exposure, compression, and resolution loss. These degradations are often spatially...
Ling-Wen Meng, Guo-Bang Ban, Jin-Tong Ma et al.· Applied Sciences· 0 citations
Through unified training, the P3-WAM model outperforms specialized baselines on 3D hand reconstruction and action-conditioned video prediction across four datasets, while producing goal-directed manipulation videos with stronger instruction adherence and action progression than video-generation baselines.
Zi-Rui Wu, Chen Guo, Tian-Jian Jiang et al.· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.