Siamese Networks for Weakly Supervised Human Activity Recognition

TL;DR

This paper introduces a siamese network-based model for human activity recognition that learns from pairwise similarity data, reducing the need for explicit labels and enabling effective clustering and recognition.

Contribution

The paper proposes a novel siamese network architecture trained with similarity information, allowing weakly supervised human activity recognition without explicit labels.

Findings

Effective in clustering and recognizing continuous human activities

Works well across multiple datasets

Reduces reliance on labeled data

Abstract

Deep learning has been successfully applied to human activity recognition. However, training deep neural networks requires explicitly labeled data which is difficult to acquire. In this paper, we present a model with multiple siamese networks that are trained by using only the information about the similarity between pairs of data samples without knowing the explicit labels. The trained model maps the activity data samples into fixed size representation vectors such that the distance between the vectors in the representation space approximates the similarity of the data samples in the input space. Thus, the trained model can work as a metric for a wide range of different clustering algorithms. The training process minimizes a similarity loss function that forces the distance metric to be small for pairs of samples from the same kind of activity, and large for pairs of samples from different kinds of activities. We evaluate the model on three datasets to verify its effectiveness in segmentation and recognition of continuous human activity sequences.