TAM: Task-Aware Memory Distillation for Efficient Spatiotemporal Prediction

Key points

• We propose TAM, a Task-Aware Memory Distillation framework that organizes a frozen teacher's knowledge into a bounded, retrievable history.1
• Memory entries encode latent features, forecast changes, or flow residuals, while task-specific selection rules identify relevant historical references.1
• We evaluate TAM on video prediction, weather forecasting, and traffic flow prediction across multiple teacher-student configurations.1
• These results demonstrate the utility of historical teacher supervision across distinct forecasting tasks without additional student inference cost.1

Sources (1)

• [1]TAM: Task-Aware Memory Distillation for Efficient Spatiotemporal Prediction
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 09:58 AM

  Knowledge distillation enables efficient spatiotemporal prediction by transferring knowledge from an accurate teacher to a compact student.

  We propose TAM, a Task-Aware Memory Distillation framework that organizes a frozen teacher's knowledge into a bounded, retrievable history.

Extractive summary: sentences quoted from the sources.