CORTEXA
← Browse

Ioannis Katsavounidis

2 papers indexed

arxiveess.IVcs.CV2026-06-28

A Self-Supervised Learning Framework for Video Encoding Complexity Clustering

Krishna Srikar Durbha, Hassene Tmar, Ping-Hao Wu, Ioannis Katsavounidis, Alan C. Bovik

Adaptive video streaming is a widely used technique for delivering video content over the internet. One of the key challenges is determining the optimal encoding settings for each video, which can vary significantly based on its content and characteristics. In this paper, we prop…

View free PDFSource page
arxivcs.CVeess.IV2026-06-28

Spatially Localized Image Degradation Embeddings for Image Quality Assessment

Krishna Srikar Durbha, Hassene Tmar, Ping-Hao Wu, Ioannis Katsavounidis, Alan C. Bovik

Self-supervised learning (SSL) currently drives state-of-the-art performance in no-reference image quality assessment (NR-IQA). However, standard SSL pipelines uniformly apply synthetic distortions across the entire image field, which can limit their sensitivity to spatially loca…

View free PDFSource page