Mostrar el registro sencillo del ítem
| dc.contributor | Radeva, Petia | |
| dc.creator | Herruzo Sánchez, Pedro | |
| dc.date | 2017-02-22T09:31:04Z | |
| dc.date | 2017-02-22T09:31:04Z | |
| dc.date | 2016-06-30 | |
| dc.date.accessioned | 2024-12-16T10:23:51Z | |
| dc.date.available | 2024-12-16T10:23:51Z | |
| dc.identifier | http://hdl.handle.net/2445/107243 | |
| dc.identifier.uri | http://fima-docencia.ub.edu:8080/xmlui/handle/123456789/16682 | |
| dc.description | Treballs Finals de Grau d'Enginyeria Informàtica, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2016, Director: Petia Radeva | |
| dc.description | Nowadays, we can find several diseases related with the unhealthy diet habits of the population, such as diabetes, obesity, anemia, bulimia and anorexia. In many cases, it is related with the food consumption of the people. Mediterranean diet is scientifically known as a healthy diet that helps to prevent those and other food problems. In particular, our work focuses on the recognition of Mediterranean food and dishes. It is part of a wider project that analyses the daily habits of users with wearable cameras, within the topic of Lifelogging. It appears as an objective tool for the analysis of the patient’s behavior, allowing specialist to discover patterns and understand user’s lifestyle to find unhealthy food patterns. With the aim to automatic recognize a complete diet, we introduce a challenging multilabeled dataset related to Mediterranean diet called FoodCAT. The first kind of labels contains 115 food classes with an average of 400 images per dish, and the second one is composed by 12 food categories with an average of 3800 pictures per class. This dataset will serve as a basis for the development of automatic diet tracking problems. Deep learning and more specifically Convolutional Neural Networks (CNNs), are actually the technologies with the state-of-the-art recognizing food automatically. In our work, we adapt the best, so far, CNNs architectures for image classification, to our objective into the diet tracking. Recognizing food categories, we achieved the highest accuracies top-1 with 72.29%, and top-5 with 97.07%. In a complete diet tracking recognizing dishes from Mediterranean diet, enlarged with the Food-101 dataset, we achieve the highest accuracies top-1 with 68.07%, and top-5 with 89.53%, for a total of 115+101 food classes. | |
| dc.format | 68 p. | |
| dc.format | application/pdf | |
| dc.language | eng | |
| dc.rights | memòria: cc-by-nc-sa (c) Pedro Herruzo Sánchez | |
| dc.rights | http://creativecommons.org/licenses/by-sa/3.0/es | |
| dc.rights | info:eu-repo/semantics/openAccess | |
| dc.source | Treballs Finals de Grau (TFG) - Enginyeria Informàtica | |
| dc.subject | Xarxes neuronals (Informàtica) | |
| dc.subject | Reconeixement de formes (Informàtica) | |
| dc.subject | Programari | |
| dc.subject | Treballs de fi de grau | |
| dc.subject | Visió per ordinador | |
| dc.subject | Cuina mediterrània | |
| dc.subject | Aprenentatge automàtic | |
| dc.subject | Neural networks (Computer science) | |
| dc.subject | Pattern recognition systems | |
| dc.subject | Computer software | |
| dc.subject | Bachelor's theses | |
| dc.subject | Computer vision | |
| dc.subject | Mediterranean cooking | |
| dc.subject | Machine learning | |
| dc.title | Can a CNN recognize mediterranean diet? | |
| dc.type | info:eu-repo/semantics/bachelorThesis |
| Ficheros | Tamaño | Formato | Ver |
|---|---|---|---|
|
No hay ficheros asociados a este ítem. |
|||