3 accepted papers
We propose a novel inference-time out-of-domain (OOD) detection algorithm for specialized large language models (LLMs). Despite achieving state-of-the-art performance on in-domain tasks through fine-tuning, specialized LLMs remain vulnerable to incorrect or unreliable outputs when presented with OOD
We address the problem of video question answering (video QA) with temporal grounding in a weakly supervised setup, without any temporal annotations. Given a video and a question, we generate an open-ended answer grounded with the start and end time. For this task, we propose TOGA: a vision-language…
Social robots are increasingly utilized across domains like education, healthcare, and elderly care. Research has examined how the robot's postures impact social perceptions, but few studies have investigated the effects of robots' bodily movements with dynamics gait variations on attributed authori