publications

For a list of my featured publications, please scroll down the home page.

Featured publications are highlighted in red.

2026

CVPR 2026 EgoVis Workshop

Action100M: A Large-scale Video Action Dataset

Delong Chen, Tejaswi Kasarla, Yejin Bang, Mustafa Shukor, Willy Chung, Jade Yu, Allen Bolourchi, Théo Moutakanni, Pascale Fung

project page paper dataset

IJCV 2026

Balanced Hyperbolic Embeddings are Natural Out-of-Distribution Detectors

Tejaswi Kasarla, Max van Spengler, Pascal Mettes

project page paper

AISTATS 2026

Hyperbolic Learning with Supervision from any Granularity

Mina Ghadimi Atigh, Max van Spengler, Teng Long, Melika Ayoughi, Tejaswi Kasarla, Pascal Mettes

paper

ICLR 2026

VL-JEPA: Joint Embedding Predictive Architecture for Vision-language

Delong Chen*, Mustafa Shukor*, Théo Moutakanni*, Willy Chung*, Jade Yu, Tejaswi Kasarla, Allen Bolourchi, Yann LeCun, Pascale Fung

paper

ICLR 2026

OrthoRF: Exploring Orthogonality in Object-Centric Representations

Despoina Touska, Bastiaan Onne Fagginger Auer, Alexandru Onose, Tejaswi Kasarla, Luis Armando Pérez Rey, Maximilian Lipp, Lyubov Amitonova, Martin R. Oswald, Pascal Cerfontaine

paper

2025

CVPR 2025 · ★ Highlight

Hyperbolic Safety-Aware Vision-Language Models

Tobia Poppi*, Tejaswi Kasarla*, Pascal Mettes, Lorenzo Baraldi, Rita Cucchiara

project page paper code model

ECCV 2025 Beyond Euclidean Workshop · ★ Oral

HierVision: Standardized and Reproducible Hierarchical Sources for Vision Datasets

Tejaswi Kasarla, Ruthu Hulikal Rooparaghunath, Stefano D’Arrigo, Gowreesh Mago, Abhishek Jha, Melika Ayoughi, Swasti Shreya Mishra, Ana Manzano Rodríguez, Teng Long, Mina Ghadimi Atigh, Max van Spengler, Pascal Mettes

project page paper code

2024

ECCV 2024 Beyond Euclidean Workshop

Maximally Separated Active Learning

Tejaswi Kasarla, Abhishek Jha, Faye Tervoort, Rita Cucchiara, Pascal Mettes

project page paper code

ICRA 2024 Workshop on Resilient Off-road Autonomy

Lightweight Uncertainty Quantification with Simplex Semantic Segmentation for Terrain Traversability

Judith Dijk, Gertjan J. Burghouts, Kapil D. Katyal, Bryanna Y. Yeh, Craig T. Knuth, Ella Fokkinga, Tejaswi Kasarla, Pascal Mettes

paper

2022

NeurIPS 2022 · ★ Oral
Also a contributed talk at WiML Workshop @ NeurIPS 2022

Maximum Class Separation as Inductive Bias in One Matrix

Tejaswi Kasarla, Gertjan J. Burghouts, Max van Spengler, Elise van der Pol, Rita Cucchiara, Pascal Mettes

project page paper code poster media

2019

WACV 2019

Region-Based Active Learning for Efficient Labelling in Semantic Segmentation

Tejaswi Kasarla, G Nagendar, Guruprasad Hegde, Vineeth N. Balasubramanian, C.V. Jawahar

project page paper supplementary poster