|
|
Deep Neural Networks: Embedded System Implementation
Matěj, Aleš ; Šimek, Václav (referee) ; Mrázek, Vojtěch (advisor)
The goal of this thesis is to firstly design and implement an application for embeddedsystems which will classify MNIST numbers and secondly optimize energy and memoryrequirements of this network. The basics of neural networks, Cortex-M processor cores andembedded devices are described in the theoretical part. Followed by implementation details.Networks learning is done with Python and Theano library on a PC. The network is thenconverted to C for a board STM32F429 Discovery. Last part consist of network optimization,which focuses on convolution, dot product and number representation of weights and biasesof the network.
|
|
|
Image Database Query by Example
Dobrotka, Matúš ; Hradiš, Michal (referee) ; Veľas, Martin (advisor)
This thesis deals with content-based image retrieval. The objective of the thesis is to develop an application, which will compare different approaches of image retrieval. First basic approach consists of keypoints detection, local features extraction and creating a visual vocabulary by clustering algorithm - k-means. Using this visual vocabulary is computed histogram of occurrence count of visual words - Bag of Words (BoW), which globally represents an image. After applying an appropriate metrics, it follows finding similar images. Second approach uses deep convolutional neural networks (DCNN) to extract feature vectors. These vectors are used to create a visual vocabulary, which is used to calculate BoW. Next procedure is then similar to the first approach. Third approach uses extracted vectors from DCNN as BoW vectors. It is followed by applying an appropriate metrics and finding similar images. The conclusion describes mentioned approaches, experiments and the final evaluation.
|
|
|
Differentiable Depth Estimation for Bin Picking
Černý, Marek ; Klusáček, David (advisor) ; Šikudová, Elena (referee)
The goal of this thesis was to investigate the neural 3D surface reconstruction from multiple views with the intent to use the resulting depth maps for bin picking. Survey of papers from 2014 to 2018 showed that none of the state of the art methods would be used to control a robot arm in our setup. Therefore we decided to create our low-level neural approach which we called the EmfNet. The network is based on a pyramidal resolution refining approach. At each pyramid's layer, there are three separate networks that take part in the computation. Each of them has a definite goal, which gives us almost complete understanding of what is going on inside the network. The EmfNet model was partially usable, but we nevertheless extended it to EmfNet-v2. First, another measuring layer was added, which freed EmfNet from depending on an unnecessary hyperparameter. Second, we used constraints on geometry for the network not to be confused by occlusions (cases where a certain part of the surface is visible only from a single camera). Both networks were implemented and tested on a corpus that was created as a part of this thesis. A corpus containing rendered as well as real data. The process of correspondence pairing inside the network can be observed using the visualization tool. We designed a way how to use a robotic arm...
|
|
|
Differentiable Depth Estimation for Bin Picking
Černý, Marek ; Klusáček, David (advisor) ; Šikudová, Elena (referee)
The goal of this thesis was to investigate the neural 3D surface reconstruction from multiple views with the intent to use the resulting depth maps for bin picking. Survey of papers from 2014 to 2018 showed that none of the state of the art methods would be used to control a robot arm in our setup. Therefore we decided to create our low-level neural approach which we called the EmfNet. The network is based on a pyramidal resolution refining approach. At each pyramid's layer, there are three separate networks that take part in the computation. Each of them has a definite goal, which gives us almost complete understanding of what is going on inside the network. The EmfNet model was partially usable, but we nevertheless extended it to EmfNet-v2. First, another measuring layer was added, which freed EmfNet from depending on an unnecessary hyperparameter. Second, we used constraints on geometry for the network not to be confused by occlusions (cases where a certain part of the surface is visible only from a single camera). Both networks were implemented and tested on a corpus that was created as a part of this thesis. A corpus containing rendered as well as real data. The process of correspondence pairing inside the network can be observed using the visualization tool. We designed a way how to use a robotic arm...
|
|
|
Deep Neural Networks: Embedded System Implementation
Matěj, Aleš ; Šimek, Václav (referee) ; Mrázek, Vojtěch (advisor)
The goal of this thesis is to firstly design and implement an application for embeddedsystems which will classify MNIST numbers and secondly optimize energy and memoryrequirements of this network. The basics of neural networks, Cortex-M processor cores andembedded devices are described in the theoretical part. Followed by implementation details.Networks learning is done with Python and Theano library on a PC. The network is thenconverted to C for a board STM32F429 Discovery. Last part consist of network optimization,which focuses on convolution, dot product and number representation of weights and biasesof the network.
|
|
|
Image Database Query by Example
Dobrotka, Matúš ; Hradiš, Michal (referee) ; Veľas, Martin (advisor)
This thesis deals with content-based image retrieval. The objective of the thesis is to develop an application, which will compare different approaches of image retrieval. First basic approach consists of keypoints detection, local features extraction and creating a visual vocabulary by clustering algorithm - k-means. Using this visual vocabulary is computed histogram of occurrence count of visual words - Bag of Words (BoW), which globally represents an image. After applying an appropriate metrics, it follows finding similar images. Second approach uses deep convolutional neural networks (DCNN) to extract feature vectors. These vectors are used to create a visual vocabulary, which is used to calculate BoW. Next procedure is then similar to the first approach. Third approach uses extracted vectors from DCNN as BoW vectors. It is followed by applying an appropriate metrics and finding similar images. The conclusion describes mentioned approaches, experiments and the final evaluation.
|
guest ::
