Vikram Bhat is a Undergraduate Research Assistant, ECE.
vgbhat@andrew.cmu.edu
Latest Research
Co-Me: Confidence-Guided Token Merging for Visual Geometry Transformer
Co-Me is an acceleration mechanism that uses a lightweight confidence predictor to selectively me...
Unified Spherical Frontend: Learning Rotation-Equivariant Representations of Spherical Images from Any Camera
A distortion-free lens-agnostic rotation-equivariant vision framework for modern perception
RADSeg: Unleashing Parameter and Compute Efficient Zero-Shot Open-Vocabulary Segmentation Using Agglomerative Models
RADSeg is a dense, language-aligned feature encoder that enables low-parameter, low-latency open-...