# Temporal Convolutional Network

The paper proposes WAND, a reinforcement learning framework for quadrotor navigation in dense obstacle fields under time-varying wind: a Temporal Convolutional Network estimates wind-induced disturbance acceleration from proprioceptive histories, the estimate is injected into the policy through a zero-initialized residual module (WindAdapter) and reused for low-level feedforward compensation, improving observed success rate by 8.3 percentage points on average over feedforward compensation alone across 12 wind-disturbed simulation settings and succeeding in 18 of 20 indoor fan-induced flight trials.

The paper proposes WAND, a reinforcement learning framework for quadrotor navigation in dense obstacle fields under time-varying wind: a Temporal Convolutional Network estimates wind-induced disturbance acceleration from proprioceptive histories, the estimate is injected into the policy through a zero-initialized residual module (WindAdapter) and reused for low-level feedforward compensation, improving observed success rate by 8.3 percentage points on average over feedforward compensation alone across 12 wind-disturbed simulation settings and succeeding in 18 of 20 indoor fan-induced flight trials.

Starting from a previously trained temporal convolutional network sperm whale click-train detector, the study compared four transfer approaches across four sources (BAL, CS, ICE, MED)—cross-dataset baseline evaluation, training from scratch on only 500 target recordings, pretraining with random fine-tuning, and pretraining with active (uncertainty-based) fine-tuning—and found that pretrained models dropped in performance on unseen data but that fine-tuning with 500 target recordings effectively mitigated the drop, with active fine-tuning consistently outperforming the other approaches although its gain over random fine-tuning was marginal.

Starting from a previously trained temporal convolutional network sperm whale click-train detector, the study compared four transfer approaches across four sources (BAL, CS, ICE, MED)—cross-dataset baseline evaluation, training from scratch on only 500 target recordings, pretraining with random fine-tuning, and pretraining with active (uncertainty-based) fine-tuning—and found that pretrained models dropped in performance on unseen data but that fine-tuning with 500 target recordings effectively mitigated the drop, with active fine-tuning consistently outperforming the other approaches although its gain over random fine-tuning was marginal.