Deep set conditioned latent representations for action recognition

TL;DR

This paper introduces deep set conditioned latent representations for improved multi-label video action recognition, leveraging relational reasoning and set-based latent features to better identify atomic and composite actions.

Contribution

It proposes SCI3D, a two-stream relational network utilizing latent set representations and reasoning to enhance action recognition accuracy.

Findings

Achieved 1.49% higher mAP in atomic action recognition.

Achieved 17.57% higher mAP in composite action recognition.

Demonstrated benefits of relational inductive biases and set-based latent representations.

Abstract

In recent years multi-label, multi-class video action recognition has gained significant popularity. While reasoning over temporally connected atomic actions is mundane for intelligent species, standard artificial neural networks (ANN) still struggle to classify them. In the real world, atomic actions often temporally connect to form more complex composite actions. The challenge lies in recognising composite action of varying durations while other distinct composite or atomic actions occur in the background. Drawing upon the success of relational networks, we propose methods that learn to reason over the semantic concept of objects and actions. We empirically show how ANNs benefit from pretraining, relational inductive biases and unordered set-based latent representations. In this paper we propose deep set conditioned I3D (SCI3D), a two stream relational network that employs latent representation of state and visual representation for reasoning over events and actions. They learn to reason about temporally connected actions in order to identify all of them in the video. The proposed method achieves an improvement of around 1.49% mAP in atomic action recognition and 17.57% mAP in composite action recognition, over a I3D-NL baseline, on the CATER dataset.