Counting People Based on Linear, Weighted, and Local Random Forests

Abstract

Recently, many works have been published for counting people. However, when being applied to real-world train station videos, they have exposed many limitations due to problems such as low resolution, heavy occlusion, various density levels and perspective distortions. In this paper, following the recent trend of regression-based density estimation, we present a linear regression approach based on local Random Forests for counting either standing or moving people on station platforms. By dividing each frame into sub-windows and extracting features with ground truth densities as well as learned weights, we perform a linear transformation for counting people to overcome the perspective problems of the existing patch-based approaches. We present improvements against several recent baselines on the UCSD dataset and a dataset of CCTV videos taken from a train station. We also show improvements in speed compared with the state-of-the-art models based on detection and Deep Learning.

A-CCNN: Adaptive CCNN for Density Estimation and Crowd Counting

“…To evaluate the performance of our A-CCNN algorithm, experiments are conducted on three challenging crowd counting datasets, i.e., the UCSD dataset [10], the UCF-CC dataset [11], and the dataset of Sydney Trains Footage (STF) [12]. Note that the first two are public benchmark datasets.…”

CNN-based Density Estimation and Crowd Counting: A Survey

“…In this subsection, we briefly introduce some special crowd counting datasets, which are only used in some certain scenarios. These datasets contain line crowd counting (LHI [137], [142], crowd sequences (PETS [138], Venice [86]), multi-sources (AHU-Crowd [139], [143], CI-ISR [149], Venice [86]), indoor (MICC [140], Indoor 1 [141], Indoor 2 [150]), train station (TS [144], STF [144]), subway station (Shanghai Subway Station [145]), BRT (Beijing BRT [146] 13 ), bridge (EBP [147]), airport (ZhengzhouAirport [151]), categorized [152]. The specific statistics of these datasets are listed in Tabel III.…”

A-CCNN: adaptive ccnn for density estimation and crowd counting

“…Table 4. Comparison of the MAE results results between A-CCNN and state-of-the-art crowd counting on STF [12].…”