D
Deepti Hegde
Researcher at KLE Technological University
Publications - 7
Citations - 29
Deepti Hegde is an academic researcher from KLE Technological University. The author has contributed to research in topics: Upsampling & Image formation. The author has an hindex of 1, co-authored 5 publications receiving 5 citations.
Papers
More filters
Journal ArticleDOI
Adaptive Cubic Spline Interpolation in CIELAB Color Space for Underwater Image Enhancement
TL;DR: A method to adaptively estimate a color correction curve in the CIE L*a*b* color space for single image enhancement is proposed and improved canny edge detection on underwater images through enhancement is demonstrated.
Journal ArticleDOI
CLIP goes 3D: Leveraging Prompt Tuning for Language Grounded 3D Recognition
TL;DR: In this article , a 3D encoder is trained using triplets of pointclouds, corresponding rendered 2D images, and texts using natural language supervision to align the features in a multimodal embedding space.
Proceedings ArticleDOI
Refining SfM Reconstructed Models of Indian Heritage Sites
TL;DR: A method to refine sparse point clouds of complex structures generated by Structure from Motion in order to achieve improved visual fidelity of ancient Indian heritage sites is proposed and compared with the state-of-the-art upsampling networks.
Book ChapterDOI
Relocalization of Camera in a 3D Map on Memory Restricted Devices
TL;DR: In this article, the authors proposed a system for relocalization which identifies 6 degrees of freedom camera pose and trajectory based on a single query image by using a vocabulary tree to create a visual word dictionary for quick and efficient image retrieval.
Posted Content
Uncertainty-aware Mean Teacher for Source-free Unsupervised Domain Adaptive 3D Object Detection
Deepti Hegde,Vishwanath A. Sindagi,Velat Kilic,A. Brinton Cooper,Mark A. Foster,Vishal M. Patel +5 more
TL;DR: The authors propose an uncertainty-aware mean teacher framework which implicitly filters incorrect pseudo-labels during training, which performs automatic soft-sampling of pseudo-labeled data while aligning predictions from the student and teacher networks.