Crowd Density Forecasting by Modeling Patch-based Dynamics
This work addresses the challenge of forecasting human activities in crowds for applications like mobile robots and autonomous driving, representing an incremental advancement in visual forecasting tasks.
The paper tackles the problem of predicting future crowd movements in surveillance videos by introducing a new task called crowd density forecasting and developing the patch-based density forecasting network (PDFN) to model density dynamics patch-wise, achieving effectiveness demonstrated on several public datasets compared to state-of-the-art methods.
Forecasting human activities observed in videos is a long-standing challenge in computer vision, which leads to various real-world applications such as mobile robots, autonomous driving, and assistive systems. In this work, we present a new visual forecasting task called crowd density forecasting. Given a video of a crowd captured by a surveillance camera, our goal is to predict how that crowd will move in future frames. To address this task, we have developed the patch-based density forecasting network (PDFN), which enables forecasting over a sequence of crowd density maps describing how crowded each location is in each video frame. PDFN represents a crowd density map based on spatially overlapping patches and learns density dynamics patch-wise in a compact latent space. This enables us to model diverse and complex crowd density dynamics efficiently, even when the input video involves a variable number of crowds that each move independently. Experimental results with several public datasets demonstrate the effectiveness of our approach compared with state-of-the-art forecasting methods.