ECCV 2026 论文列表

Translating continuous, noisy egocentric video streams into discrete, temporally ordered action steps is fraught with visual challenges. Heavy ego-motion, transient occlusions, and the high intra-class variability of unscripted human-object interactions cause standard framelevel online temporal models to struggle, often resulting in severe oversegmentation and structural collapse. To bridge the gap between unstable low-level perception and high-level procedural logic, we present VidParse, a online, training-free framework that treats activity understanding as a graph-constrained inference problem. Rather than relying on learned temporal filters, we dynamically identify semantic transitions using a temporal similarity matrix over manipulation-anchored features, which are extracted from frozen foundation models to prioritize foreground hand-object interactions. A beam search decoder then leverages an induced procedural task graph to explicitly enforce valid action transitions and prune impossible trajectories. By anchoring robust visual segments to hard procedural constraints, our approach preserves longrange state transitions and achieves up to a 10x improvement in complex multi-step parsing accuracy over strong online baselines, all without requiring a single gradient update.

Reliable localization of non-line-of-sight (NLOS) pedestriansis critical for safe urban autonomous driving, yet it remains highly chal-lenging in ego-dynamic outdoor environments, where ego-vehicle motionmakes radar multipath propagation complex and noisy. In this paper, wepresent a reflection-aware framework for NLOS pedestrian localizationwith a moving ego-vehicle in outdoor testbed scenarios. Our frameworkfuses front-view camera images and 2D radar point clouds to infer reflec-tion orders and reflective surface distributions in bird’s-eye-view space.It then uses physics-guided ray tracing to reconstruct distorted reflec-tion paths and localize the hidden pedestrian. We validate the frame-work in outdoor testbed scenarios under ego-dynamic conditions. The re-sults demonstrate the effectiveness of the proposed framework for NLOSpedestrian localization with a moving ego-vehicle.