Benchmarking algorithms for food localization and semantic segmentation
- PDF / 7,620,092 Bytes
- 21 Pages / 595.276 x 790.866 pts Page_size
- 42 Downloads / 193 Views
ORIGINAL ARTICLE
Benchmarking algorithms for food localization and semantic segmentation Sinem Aslan1,2,3 · Gianluigi Ciocca4 · Davide Mazzini4 · Raimondo Schettini4 Received: 24 August 2019 / Accepted: 9 June 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract The problem of food segmentation is quite challenging since food is characterized by intrinsic high intra-class variability. Also, segmentation of food images taken in-the-wild may be characterized by acquisition artifacts, and that could be problematic for the segmentation algorithms. A proper evaluating of segmentation algorithms is of paramount importance for the design and improvement of food analysis systems that can work in less-than-ideal real scenarios. In this paper, we evaluate the performance of different deep learning-based segmentation algorithms in the context of food. Due to the lack of largescale food segmentation datasets, we initially create a new dataset composed of 5000 images of 50 diverse food categories. The images are accurately annotated with pixel-wise annotations. In order to test the algorithms under different conditions, the dataset is augmented with the same images but rendered under different acquisition distortions that comprise illuminant change, JPEG compression, Gaussian noise, and Gaussian blur. The final dataset is composed of 120,000 images. Using standard benchmark measures, we conducted extensive experiments to evaluate ten state-of-the-art segmentation algorithms on two tasks: food localization and semantic food segmentation. Keywords Benchmarking · Convolutional neural network · Food localization · Food segmentation
1 Introduction
* Gianluigi Ciocca [email protected] Sinem Aslan [email protected] Davide Mazzini [email protected] Raimondo Schettini [email protected] 1
Present Address: Ca’ Foscari University of Venice, ECLT, Ca’ Bottacin, Dorsoduro 3911, Calle Crosera, 30123 Venice, Italy
2
Ca’ Foscari University of Venice, DAIS, via Torino 155, 30172 Mestre, VE, Italy
3
International Computer Institute, Ege University, 35100 Izmir, Bornova, Turkey
4
Department of Informatics, Systems and Communications, University of Milano-Bicocca, Viale Sarca 336, 20126 Milan, Italy
The problem of a healthy and balanced meal is seriously tackled by the different health agencies to reduce obesity and unbalanced nutrition. Accurate tracking of daily nutrition intake is not only conducive for people to maintain a healthy weight but also essential to treat and control food-related health problems like obesity and diabetes [4]. Conventionally, such procedure has been accomplished by exploring logs manually recorded each day, howbeit manual recording is error-prone due to several factors like the inability to estimate the food type and quantity, difficulty to provide the continuity of such a demanding task, and delayed reporting. On the other hand, computer vision algorithms have come as a remedy for a more accurate and user-friendly procedure by automatizing di
Data Loading...
