Spatiotemporal Action Recognition in Restaurant Videos

TL;DR

This paper investigates spatiotemporal action recognition in restaurant videos, focusing on small objects and rapid actions, by adapting YOLO and proposing a novel recurrent YOLO variant with convolutional LSTMs.

Contribution

It introduces a recurrent modification of YOLO with convolutional LSTMs and evaluates the effectiveness of YOWO's 3D convolutions on a unique, challenging dataset.

Findings

Recurrent YOLO with LSTMs improves action detection accuracy.

YOWO's 3D convolutions effectively capture spatiotemporal features.

The approaches handle small objects and rapid actions in restaurant videos.

Abstract

Spatiotemporal action recognition is the task of locating and classifying actions in videos. Our project applies this task to analyzing video footage of restaurant workers preparing food, for which potential applications include automated checkout and inventory management. Such videos are quite different from the standardized datasets that researchers are used to, as they involve small objects, rapid actions, and notoriously unbalanced data classes. We explore two approaches. The first approach involves the familiar object detector You Only Look Once, and another applying a recently proposed analogue for action recognition, You Only Watch Once. In the first, we design and implement a novel, recurrent modification of YOLO using convolutional LSTMs and explore the various subtleties in the training of such a network. In the second, we study the ability of YOWOs three dimensional convolutions to capture the spatiotemporal features of our unique dataset