Kevin Issac

I am an ELLIS/ELIZA Doctoral Researcher in the Visual Inference Lab at TU Darmstadt, supervised by Stefan Roth. My research focuses on computer vision and machine learning, with an emphasis on ill-posed problems such as sparse-view scene synthesis.

I obtained my Master’s degree in Visual Computing from Saarland University. During my Master’s, I was supervised by Jan Eric Lenssen at the Max Planck Institute for Informatics, Germany. Prior to that, I worked as a Research Assistant at the Spectrum Lab, Indian Institute of Science (IISc), where I conducted research in biomedical image processing under the supervision of Chandra Sekhar Seelamantula and Harish Kumar JR.

computer graphics · neural representations · geometry learning · computer vision

Email / CV / Github / Scholar

News

Invited to be a CVPR'26 reviewer.
Spurfies accepted to 3DV'25 as an oral and nominated as an award candidate.
Presented latentSplat at ECCV'24 on Oct 3 2024.
latentSplat accepted to ECCV'24.
Successfully defended my Master's Thesis on April 9 2024.

Publications

	Spurfies: Sparse Surface Reconstruction using Local Geometry Priors Kevin Raj, Christopher Wewer, Raza Yunus, Eddy Ilg, Jan Eric Lenssen International Conference on 3D Vision (3DV), Singapore, 2025 (Oral Presentation) (Award Candidate) Paper / Project Website / BibTeX
	latentSplat: Autoencoding Variational Gaussians for Fast Generalizable 3D Reconstruction Christopher Wewer, Kevin Raj, Eddy Ilg, Bernt Schiele, Jan Eric Lenssen European Conference on Computer Vision (ECCV), Milan, Italy., 2024 Paper / Project Website / BibTeX
	Automatic Classification of Artery-Vein from a Single Wavelength Fundus Images Kevin Raj, Aniketh Manjunath, J.R.H. Kumar, Chandra S. Seelamantula IEEE International Symposium on Biomedical Imaging (ISBI), Iowa, USA., 2020 Paper / Supplementary / BibTeX
	A Structure Tensor based Voronoi Decomposition Technique for Optic Cup Segmentation Kevin Raj, J.R.H Kumar, S. Jois, S. Harsha, Chandra S. Seelamantula IEEE International Conference on Image Processing (ICIP), Taipei, Taiwan., 2019 (Oral Presentation) Paper / Supplementary / BibTeX
	Automatic Segmentation of Common Carotid Artery in Longitudinal Mode Ultrasound Images Using Active Oblongs J.R.H. Kumar, K. Teotia, Kevin Raj, A. Jasbon, K.V. Rajagopal, Chandra S. Seelamantula IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Brighton, UK., 2019 Paper / BibTeX

Projects

	Data Collection. Project Website
	Computer Graphics-1: Rendering Competition. Project Website
	Few-shot Semantic Segmentation of Wireless Capsule Endoscopy Images Paper / Code

Thanks to Jon Barron for making his template open source.