We are excited to announce that two papers from our research group in the Department of Computer Science and Engineering at IIT Dharwad have been accepted for publication at the prestigious A* conference in computer vision, the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV-2025). The conference will take place in Tucson, Arizona, from February 28 to March 3, 2025, where the papers will be presented.
Paper 1: Title: "When Visual State Space Model Meets Backdoor Attack"
Authors: Sankalp Nagaonkar, Achyut Mani Tripathi, Ashish Mishra
This work presents the first-ever imperceptible backdoor attack against the Visual State Space model (VMamba). BTP Thesis of Mr. Sankalp Nagaonkar.
Paper 2: Title: "Bandit Based Attention Mechanism in Vision Transformer"
Authors: Raghuram Bharadwaj, Prabhuchandran K J, Achyut Mani Tripathi
This work explores how reinforcement learning (RL) can be utilized to enhance the computational efficiency of the Vision Transformer (ViT) model.
The above works have been supervised by a group that includes Dr. Achyut Mani Tripathi (IIT Dharwad), Dr. Prabu Chandran K J (IIT Dharwad), Dr. Raghuram Bharadwaj (IIIT-Bangalore), and Dr. Ashish Mishra (ML/AI Data Scientist at HPE, Bangalore). Congratulations to all the Co-Authors!