YOUKU-VSRE 2019 Video Enhancement and Super-Resolution Challenge Announcement

Video enhancement and super‑resolution are core algorithms in computer vision, aiming to restore degraded video content and improve its clarity. Beyond academic interest, these technologies have significant industrial relevance, especially for improving the quality and sharpness of early film footage.

The YOUKU‑VSRE 2019 challenge provides a platform to showcase state‑of‑the‑art video enhancement and super‑resolution methods. The competition supplies the currently largest and most diverse dataset, covering various content categories, noise models, and difficulty levels. The data generation fully simulates real‑world noise patterns, enabling researchers to refine algorithms for practical scenarios.

Challenge task: participants train models on the provided training set and then predict high‑resolution videos from low‑resolution inputs.

Competition format: three stages – preliminary (May 5 – June 18), semi‑final (June 25 – August 10), and final (September, date TBD). In the first two stages, participants submit high‑definition results via a server; top‑ranking teams advance. The final stage is an on‑site presentation, and the final score combines competition results with the presentation assessment.

Prize settings: Champion – 1 team, ¥100,000 and certificate; Runner‑up – 1 team, ¥60,000 and certificate; Third place – 1 team, ¥40,000 and certificate; Geek Award – 3 teams, ¥10,000 each and certificates. Additionally, the top 10 teams from the semi‑final receive a fast‑track recruitment channel at Alibaba (Youku).

For more detailed information, please visit the official competition website.