Litcius/Paper detail

Visual-Linguistic Representation Learning with Deep Cross-Modality Fusion for Referring Multi-Object Tracking

Wenyan He, Yajun Jian, Lu Yang, Hanzi Wang

202410 citationsDOI

Abstract

Referring multi-object tracking is a new rising research topic that aims at detecting and tracking the referred objects in a video sequence based on a natural language expression. Compared with traditional multi-object tracking, this setting guides object tracking with high-level semantic information, which may bring more flexible and robust tracking performance in practical scenarios. However, existing methods perform cross-modal fusion in only one phase. The limitation of visual-linguistic representation is prone to causing visionlanguage mismatching and producing poor tracking results. To effectively fuse vision and language modalities, we propose DeepRMOT with deep cross-modality fusion, including an enhanced early-fusion module, a bidirectional crossmodality encoder, and a cross-modality decoder. Therefore, DeepRMOT can boost object detection and data association by the enhanced visual-linguistic representation. Extensive experiments on the Refer-KITTI demonstrate the effectiveness of our method.

Topics & Concepts

Computer scienceArtificial intelligenceModality (human–computer interaction)Video trackingComputer visionRepresentation (politics)Object (grammar)Tracking (education)Eye trackingObject detectionEncoderFuse (electrical)Natural language processingPattern recognition (psychology)EngineeringPolitical sciencePsychologyOperating systemPedagogyPoliticsLawElectrical engineeringVideo Surveillance and Tracking MethodsMultimodal Machine Learning ApplicationsAdvanced Image and Video Retrieval Techniques
Visual-Linguistic Representation Learning with Deep Cross-Modality Fusion for Referring Multi-Object Tracking | Litcius