Propostas para Dissertação

Mestrados no Departamento de Informática

Consultar ficha completa de uma proposta de dissertação

proponente: Andre Leite Ferreira
instituição/empresa: Bosch Car Multimedia
tema/título: Platform for video processing from Far Infrared cameras (FAR) using NVIDIA CUDA technology.
área científica: Data mining/Deep Learning
local: Bosch Car Multimedia S.A.
curso de mestrado: Mestrado Integrado em Engenharia Informática

In a context of autonomous driving, perceiving the environment of a vehicle requires the ability to
detect the presence of objects outside or inside the vehicle. This information when captured is
relevant for defining vehicle intelligent responses to events occurring in vehicle surroundings and
interior. The project aims to define and implement a platform for image collection and pro-cessing
based on NVIDIA Tegra TX2 hardware. Project Objetives 1st phase - 12 weeks 1- Setup system for data collection of FLIR cameras on a real vehicle. 2- Literature review on the topic of image processing using GPU architectures. Problem definition
and research plan. 3- Thesis plan writing. 2nd phase - 35 weeks 1- Configure and implement ‘standard’ classification algorithms (using deep learning frameworks
like Caffe, CNTK, TensorFlow, Theano and Torch) on NVIDIA Tegra TX2 boards (using CUDA toolkit). 2- Benchmark performance of different implementations. 3- Thesis writing Relevant Literature - Deep Learning by Ian Goodfellow, Yoshua Bengio, Aaron Courville - Hands-On Machine Learning with Scikit-Learn and TensorFlow: Concepts, Tools, and Techniques to
Build Intelligent Systems 1st Edition by Aurélien Géron - NVIDIA Tegra - - CUDA toolkit -