A Novel Prediction Method for Early Recognition of Global Human Behaviour in Image Sequences

Abstract

Human behaviour recognition has been, and still remains, a challenging problem that involves different areas of computational intelligence. The automated understanding of people activities from video sequences is an open research topic in which the computer vision and pattern recognition areas have made big efforts. In this paper, the problem is studied from a prediction point of view. We propose a novel method able to early detect behaviour using a small portion of the input, in addition to the capabilities of it to predict behaviour from new inputs. Specifically, we propose a predictive method based on a simple representation of trajectories of a person in the scene which allows a high level understanding of the global human behaviour. The representation of the trajectory is used as a descriptor of the activity of the individual. The descriptors are used as a cue of a classification stage for pattern recognition purposes. Classifiers are trained using the trajectory representation of the complete sequence. However, partial sequences are processed to evaluate the early prediction capabilities having a specific observation time of the scene. The experiments have been carried out using the three different dataset of the CAVIAR database taken into account the behaviour of an individual. Additionally, different classic classifiers have been used for experimentation in order to evaluate the robustness of the proposal. Results confirm the high accuracy of the proposal on the early recognition of people behaviours.

High-Level Video Event Modeling, Recognition, and Reasoning via Petri Net

“…We also conduct a series of experiments to compare the sensitivity, specificity, and accuracy of Petri net based methods and non-Petri net based method [20]. The definitions of the three performance indicator can be found in [20].…”

“…By semantic reasoning, the system is able to connect the simple activities into more complex activities. Azorin-Lopez et al [20] propose a predictive method based on a simple representation of trajectories of a person in the scene. It allows a high level understanding of the global human behavior.…”

Human Interaction Anticipation by Combining Deep Features and Transformed Optical Flow Components

“…Interactions are represented with integral bag-of-words and dynamic bag-ofwords. Lopez et al [13] predicted a portion of trajectories by using a simple trajectory-based representation named as activity description vector (ADV). The ADV is composed of frequency and the four directions of each point i.e.…”

Constrained self-organizing feature map to preserve feature extraction topology

“…The ADV is a trajectorybased feature presented in previous works to describe global human behaviour [5] and was used as the input of an early prediction method [7,6]. For the sake of completeness, the ADV is presented but we refer you to [5] in order to obtain further details about its calculation.…”

“…In this case, the nD-SOM-PINT uses as input an ADV collection extracted from a sequence of two-dimensional spaces, a sequence of images. In previous works [5,7,6], a vector concatenating all ADV i,j was used as the collection of ADVs describing a person trajectory. Hence, for a scenario space of uxv, split in mxn cells, the ADV will contain mxnx5 elements as:…”