Interactive Multi-Turn Retrieval for Health Videos
For health professionals and patients seeking instructional videos, this work provides a scalable benchmark and retrieval method that handles vague initial queries refined through follow-up constraints.
The paper addresses the brittleness of single-turn retrieval for health instructional videos by introducing interactive multi-turn retrieval. It constructs a benchmark (MHVRC) and proposes a two-stage framework (DATR) that achieves consistent gains over baselines, with user studies showing multi-turn queries better capture fine-grained procedural semantics.
The growing availability of health-related instructional videos creates new opportunities for clinical training, patient rehabilitation, and health education, yet existing retrieval systems remain largely single-turn: a user submits one query and receives one ranked list. This interaction is brittle in health scenarios, where information needs are often vague at first and become clinically meaningful only after follow-up constraints such as posture, hand placement, contraindications, equipment, or patient condition are specified. We introduce interactive multi-turn semantic retrieval for health videos and construct MHVRC, a Multi-Turn Health Video Retrieval Corpus, by combining video-grounded descriptions from VideoChat-Flash with query refinements generated by DeepSeek. We further propose DATR, a Dialogue-Aware Two-Stage Retrieval framework. DATR first performs efficient coarse retrieval with a CLIP-style dual encoder and sparse frame sampling, then re-ranks the top candidates through multi-turn query fusion and a lightweight cross-encoder scoring module. Experiments on MHVRC show consistent gains over strong text-video retrieval baselines, while user studies indicate that refined multi-turn queries better capture fine-grained procedural semantics than single-turn annotations. The work establishes a benchmark and a scalable technical recipe for interactive health video retrieval.