The Activis works to provide a practical help for visual impaired people, developing an Android application to navigate in indoor environment. The need for an Object Detection model for this purpose is clear. In this thesis work are proposed and analyzed the performances of two models: SSD-Lite with Mobilenet V2 and Tiny-DSOD. It is also proposed a new dataset called Office dataset, useful to test the two algorithms.
Deep Learning Networks for Real-time Object Detection on Mobile Devices
Tramontano, Andrea Gaetano
2019/2020
Abstract
The Activis works to provide a practical help for visual impaired people, developing an Android application to navigate in indoor environment. The need for an Object Detection model for this purpose is clear. In this thesis work are proposed and analyzed the performances of two models: SSD-Lite with Mobilenet V2 and Tiny-DSOD. It is also proposed a new dataset called Office dataset, useful to test the two algorithms.File in questo prodotto:
| File | Dimensione | Formato | |
|---|---|---|---|
|
andrea_gaetano_tramontano_tesi.pdf
accesso aperto
Dimensione
2.17 MB
Formato
Adobe PDF
|
2.17 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
Utilizza questo identificativo per citare o creare un link a questo documento:
https://hdl.handle.net/20.500.12608/24599