No Trust Issues Here: A Technical Report on the Winning Solutions for the Rayan AI Contest

TL;DR

This report details winning solutions for the Rayan AI Contest, showcasing advances in compositional image retrieval, zero-shot anomaly detection, and backdoored model detection with high accuracy and top rankings.

Contribution

We present novel methods for three AI challenges, achieving top placements and demonstrating effectiveness in retrieval, anomaly detection, and model security.

Findings

Achieved 95.38% accuracy in image retrieval

Secured second place with 73.14% in anomaly detection

Detected backdoored models with 78% accuracy

Abstract

This report presents solutions to three machine learning challenges developed as part of the Rayan AI Contest: compositional image retrieval, zero-shot anomaly detection, and backdoored model detection. In compositional image retrieval, we developed a system that processes visual and textual inputs to retrieve relevant images, achieving 95.38% accuracy and ranking first with a clear margin over the second team. For zero-shot anomaly detection, we designed a model that identifies and localizes anomalies in images without prior exposure to abnormal examples, securing second place with a 73.14% score. In the backdoored model detection task, we proposed a method to detect hidden backdoor triggers in neural networks, reaching an accuracy of 78%, which placed our approach in second place. These results demonstrate the effectiveness of our methods in addressing key challenges related to retrieval, anomaly detection, and model security, with implications for real-world applications in industries such as healthcare, manufacturing, and cybersecurity. Code for all solutions is available online (https://github.com/safinal/rayan-ai-contest-solutions).