The field of computer vision and sensing is moving towards more innovative and effective methods for tasks such as gaze estimation, depth estimation, and object detection. Researchers are exploring the use of deep learning algorithms and novel network designs to improve the accuracy and robustness of these systems. One notable trend is the development of methods that can operate effectively in realistic mobile usage conditions, such as varying lighting conditions and device types. Another area of focus is the integration of historical context and memory mechanisms into detection architectures to improve their ability to handle challenging scenes. Noteworthy papers in this area include:

• Evaluating Sensitivity Parameters in Smartphone-Based Gaze Estimation, which demonstrates the potential of appearance-based gaze estimation methods.
• DiFuse-Net, which introduces a novel modality decoupled network design for disentangled RGB and dual-pixel depth estimation.
• MonoVQD, which proposes a novel framework for monocular 3D object detection with variational query denoising and self-distillation.
• Retrospective Memory for Camouflaged Object Detection, which proposes a recall-augmented architecture for camouflaged object detection that integrates historical context into the detection process.