AI CUP 2026 Fall Contest on Solar Panel Defect Detection
Guided by the Department of Information and Technology Education, Ministry of Education, and administered by the Ministry of Education AI Contest and Annotation Data Collection Project Office. Students in this program are welcome to form a team.
The official contest name is the AI CUP 2026 Fall Contest on solar panel defect detection and root cause analysis, combining few-shot learning, data generation, and multimodal vision-language models (VLM). Participants must detect defect locations from visible-light (RGB) and thermal infrared (Thermal IR) images and produce written root-cause descriptions. Registration has been extended to 23:59:59 on Monday, November 30, 2026. Please complete registration on the official AI CUP website.
The contest focuses on image-based defect detection and root-cause description, skills that are also useful in semiconductor inspection and process problem analysis.
Result submission until: 23:59:59, December 4, 2026
Private Leaderboard release: December 9, 2026
Report and code upload: December 9 to 16, 2026
Final ranking announcement: January 13, 2027
The award ceremony is tentatively scheduled for March 2027.
Cash prizes and other material awards are limited to teams whose members are all enrolled students with Republic of China student status at registration. The top 15 student teams may receive a Ministry of Education certificate after review.
Teams of 1 to 5 people may register. Enrolled students with Republic of China student status at registration (including senior high school, undergraduate, and graduate students) may form a team. Members of the public may also take part and appear in the overall ranking, but they are not eligible for cash prizes or other material awards.
This is a forwarded notice. Official rules, eligibility, and awards follow announcements by the contest organizer.
Contact: aicup2026.solardefect@gmail.com. Register at https://go.aicup.tw/.


