MACNet: Multi-scale Atrous Convolution Networks for Food Places Classification in Egocentric Photo-streams

TL;DR

This paper introduces MACNet, a deep multi-scale atrous convolution network designed to automatically recognize food places in egocentric photo-streams, aiding in understanding personal food interaction patterns.

Contribution

The paper proposes a novel deep end-to-end model using multi-scale atrous convolutions for food place recognition in egocentric images, evaluated on a new private dataset.

Findings

Achieved promising accuracy in food place classification

Demonstrated effectiveness of multi-scale atrous convolutions

Validated model on the EgoFoodPlaces dataset

Abstract

First-person (wearable) camera continually captures unscripted interactions of the camera user with objects, people, and scenes reflecting his personal and relational tendencies. One of the preferences of people is their interaction with food events. The regulation of food intake and its duration has a great importance to protect against diseases. Consequently, this work aims to develop a smart model that is able to determine the recurrences of a person on food places during a day. This model is based on a deep end-to-end model for automatic food places recognition by analyzing egocentric photo-streams. In this paper, we apply multi-scale Atrous convolution networks to extract the key features related to food places of the input images. The proposed model is evaluated on an in-house private dataset called "EgoFoodPlaces". Experimental results shows promising results of food places classification recognition in egocentric photo-streams.