The field of computer vision and multimodal processing is rapidly evolving, with a strong focus on developing innovative methods to integrate and process different types of data, such as RGB, depth, and event-based information. Recent research has emphasized the importance of effectively combining these modalities to improve performance in various applications, including object tracking, segmentation, and recognition. Furthermore, the development of adversarial attacks has become a crucial aspect of this field, as researchers seek to understand and mitigate the vulnerabilities of deep learning models to malicious inputs. Notably, the creation of novel attack frameworks and the improvement of existing ones have led to significant advancements in the field, highlighting the need for robust defense mechanisms. Noteworthy papers in this area include HDBFormer, which proposes a heterogeneous dual-branch framework for efficient RGB-D semantic segmentation, and Rethinking Target Label Conditioning in Adversarial Attacks, which introduces a 2D tensor-guided generative approach for multi-target attacks. Additionally, the paper on Human-Imperceptible Physical Adversarial Attack for NIR Face Recognition Models presents a stealthy and practical adversarial patch to attack NIR face recognition systems, demonstrating improved attack success rates compared to state-of-the-art methods.