Object Detection

Generic object detection is one of the key tasks in computer vision. Due to the evolution of deep learning algorithms, we have seen remarkable improvements in this field. Along with generic object detection, our research work focuses on its applications like firearms detection. We are working on the weakly supervised orientation aware object detection which allows to suppress the background information and propagates the meaningful foreground information. Along with this, we work on human firearm pair localization to identify the carriers. Due to the limitations of annotated data for object detection, we are also working in the field of domain adaptation for object detection to alleviate or reduce the domain gap in the source and target distributions.


Most of the recent Deep Semantic Segmentation algorithms suffer from large generalization errors, even when powerful hierarchical representation models based on convolutional neural networks have been employed. This could be attributed to…