From Where How

Egocentric 4D interaction forecasting aims to anticipate both where future interactions will occur in 3D and how the human body will move to realize them, providing an important capability for assistive robotics and human–computer interaction. Existing methods typically model interaction localization and full-body motion separately, or represent interaction locations using discrete spatial grids, leaving their continuous temporal and geometric correspondence insufficiently modeled. In addition, accurate interaction forecasting requires jointly capturing task semantics, metric localization, and short-horizon visual dynamics, while stochastic pose forecasting must balance motion diversity with structural consistency. To address these challenges, we introduce Coherent4D, a large-scale egocentric dataset for continuous 4D interaction forecasting, comprising approximately 233K samples across Cooking, Health, and Bike Repair. Each sample pairs an ordered sequence of future 3D interaction locations with temporally aligned full-body poses in a shared coordinate system, together with continuous-space evaluation metrics. Building on this formulation, we propose HIGFlow, a Hand-Interaction-Guided Residual Flow framework that models forecasting as a cascaded where-to-how process. HIGFlow first predicts continuous future interaction locations by combining semantic grounding with short-horizon visual dynamics, and then uses the predicted location sequence to condition a deterministic motion anchor and residual Flow Matching for diverse yet structurally consistent full-body motion forecasting. Extensive experiments across all three domains demonstrate consistent improvements over representative baselines on both interaction-location and pose forecasting, while ablations validate the contributions of the proposed components.