|国家预印本平台
首页|CCLSTM: Coupled Convolutional Long-Short Term Memory Network for Occupancy Flow Forecasting

CCLSTM: Coupled Convolutional Long-Short Term Memory Network for Occupancy Flow Forecasting

CCLSTM: Coupled Convolutional Long-Short Term Memory Network for Occupancy Flow Forecasting

来源:Arxiv_logoArxiv
英文摘要

Predicting future states of dynamic agents is a fundamental task in autonomous driving. An expressive representation for this purpose is Occupancy Flow Fields, which provide a scalable and unified format for modeling motion, spatial extent, and multi-modal future distributions. While recent methods have achieved strong results using this representation, they often depend on high-quality vectorized inputs, which are unavailable or difficult to generate in practice, and the use of transformer-based architectures, which are computationally intensive and costly to deploy. To address these issues, we propose \textbf{Coupled Convolutional LSTM (CCLSTM)}, a lightweight, end-to-end trainable architecture based solely on convolutional operations. Without relying on vectorized inputs or self-attention mechanisms, CCLSTM effectively captures temporal dynamics and spatial occupancy-flow correlations using a compact recurrent convolutional structure. Despite its simplicity, CCLSTM achieves state-of-the-art performance on occupancy flow metrics and, as of this submission, ranks \(1^{\text{st}}\) in all metrics on the 2024 Waymo Occupancy and Flow Prediction Challenge leaderboard.

Peter Lengyel

自动化技术、自动化技术设备计算技术、计算机技术

Peter Lengyel.CCLSTM: Coupled Convolutional Long-Short Term Memory Network for Occupancy Flow Forecasting[EB/OL].(2025-06-06)[2025-06-27].https://arxiv.org/abs/2506.06128.点此复制

评论