This paper introduces Q-Tag, the first watermarking framework for Quantum Circuit Generative Models (QCGMs). It addresses the challenge of protecting intellectual property in automatically synthesized quantum circuits by embedding ownership signals directly into the generative process using a symmetric sampling strategy and a synchronization mechanism, ensuring both high fidelity and robust watermark detection against various attacks.

This PhD thesis presents innovations to improve the efficiency of normalizing flows through new architectures and algorithms, alongside applying generative models to diverse computer vision challenges such as agricultural quality assessment, privacy-preserving autonomous driving, geological mapping, art restoration, and missing traffic sign detection.

This paper introduces C-DiffDet+, a novel object detection framework that enhances DiffusionDet by integrating global scene context with local proposal features. This approach significantly improves car damage detection, especially for challenging fine-grained damage like scratches and cracks, by leveraging context-aware fusion mechanisms.