5 papers
VL-JEPA: Joint Embedding Predictive Architecture for Vision-language
Delong Chen, Mustafa Shukor, Theo Moutakanni +7
We introduce VL-JEPA, a vision-language model built on a Joint Embedding Predictive Architecture (JEPA). Instead of autoregressively generating tokens as in classical VLMs, VL-JEPA…
Action100M: A Large-scale Video Action Dataset
Delong Chen, Tejaswi Kasarla, Yejin Bang +6
Inferring physical actions from visual observations is a fundamental capability for advancing machine intelligence in the physical world. Achieving this requires large-scale, open-…
Balanced Hyperbolic Embeddings Are Natural Out-of-Distribution Detectors
Tejaswi Kasarla, Max van Spengler, Pascal Mettes
Out-of-distribution recognition forms an important and well-studied problem in deep learning, with the goal to filter out samples that do not belong to the distribution on which a…
Hyperbolic Safety-Aware Vision-Language Models
Tobia Poppi, Tejaswi Kasarla, Pascal Mettes +2
Addressing the retrieval of unsafe content from vision-language models such as CLIP is an important step towards real-world integration. Current efforts have relied on unlearning t…
Maximally Separated Active Learning
Tejaswi Kasarla, Abhishek Jha, Faye Tervoort +2
Active Learning aims to optimize performance while minimizing annotation costs by selecting the most informative samples from an unlabelled pool. Traditional uncertainty sampling o…