Ameya Phalak

Papers

Total Citations

H-Index

About

Ameya Phalak is a computer vision researcher whose work bridges deep learning and 3D scene understanding, with a focus on indoor spatial reconstruction. His most cited contribution, "DeepPerimeter: Indoor Boundary Estimation from Posed Monocular Sequences" (2019, 9 citations), introduces a novel deep learning pipeline that infers full indoor perimeter maps—essentially exterior boundary layouts—from a sequence of posed RGB images. By integrating robust depth estimation and wall segmentation, Phalak’s method generates a boundary point cloud, enabling accurate reconstruction of room geometry without relying on expensive sensors like LiDAR. This work has implications for robotics, augmented reality, and architectural modeling, where understanding spatial constraints from simple camera inputs is critical. While his citation count reflects an emerging career, the technical novelty of DeepPerimeter—combining deep learning with geometric reasoning—marks a meaningful step toward accessible indoor mapping. Phalak’s research demonstrates how neural networks can extract structural insights from everyday imagery, offering a practical tool for applications ranging from autonomous navigation to virtual staging.