|
|
Detection of Vehicles in Image and Video
Petráš, Adam ; Zemčík, Pavel (referee) ; Špaňhel, Jakub (advisor)
This bachelor thesis is focused on vehicle detection. The thesis deals with the method of vehicle detection using convolutional neural networks, their structures and models. All scripts were implemented using python programming language with Tensorflow Object Detection API interface. The first part of this thesis was devote to the structures of popular neural networks and models of detection neural networks. The next chapter deals with the most famous frameworks that are used for machine learning. Three neural network models were selected and trained on the COD20K dataset. The result of this thesis is statistics that discuss the efficiency and performance of each model on trained dataset and compare performance without displaying video on Nvidia RTX 2060, where the performace archieved by SSD MobileNet V2 network was 300FPS and Nvidia Tegra TX2 8GB, whose performace reached almost 44FPS.
|
|
|
Restoration of audio signals damaged by quantization
Šiška, Jakub ; Rajmic, Pavel (referee) ; Záviška, Pavel (advisor)
This master’s thesis deals with the restoration of audio signals damaged by quantization. The theoretical part starts with a description of quantization and dequantization in general, few existing methods of dequantization of audio signals and theory of sparse representations of signals are also presented. The next part introduces algorithms suitable for dequantization, specifically Douglas–Rachford, Chambolle–Pock, SPADEQ and implementation of these algorithms in MATLAB application in the next chapter. In the last part of this thesis, testing of reconstructed signals using the algorithms takes place and results are evaluated by objective measures SDR, PEMO-Q, PEAQ and subjective listening test MUSHRA.
|
|
|
Practical examples of signal processing
Hanzálek, Pavel ; Smékal, Zdeněk (referee) ; Mekyska, Jiří (advisor)
The thesis focuses on the issue of signal processing. Using practical examples, it tries to show the use of individual signal processing operations from a practical point of view. For each of the selected signal processing operations, an application is created in MATLAB, including a graphical interface for easier operation. The division of the thesis is such that each chapter is first analyzed from a theoretical point of view, then it is shown using a practical demonstration of what the operation is used in practice. Individual applications are described here, mainly in terms of how they are handled and their possible results. The results of the practical part are presented in the attachment of the thesis.
|
|
|
The effect of sampling period error on frequency analysis
Pietrowicz, Daniel ; Marada, Tomáš (referee) ; Zuth, Daniel (advisor)
This thesis will be focused on signal measurement in connection with digitization and analysis inaccuracy of sampling peroid effect. Brief narration of frequency analysis and description of Matlab and his capabilities with this issue. Then creation procedure for simulation model of analog signal and his fault digitization via sampling period. Systematic comparation of this effect on inaccuracy of sampling period on frequency analysis.
|
|
|
Exploiting Approximate Arithmetic Circuits in Neural Networks Inference
Matula, Tomáš ; Mrázek, Vojtěch (referee) ; Češka, Milan (advisor)
Táto práca sa zaoberá využitím aproximovaných obvodov v neurónových sieťach so zámerom prínosu energetických úspor. K tejto téme už existujú štúdie, avšak väčšina z nich bola príliš špecifická k aplikácii alebo bola demonštrovaná v malom rozsahu. Pre dodatočné preskúmanie možností sme preto skrz netriviálne modifikácie open-source frameworku TensorFlow vytvorili platformu umožňujúcu simulovať používanie approximovaných obvodov na populárnych a robustných neurónových sieťach ako Inception alebo MobileNet. Bodom záujmu bolo nahradenie väčšiny výpočtovo náročných častí konvolučných neurónových sietí, ktorými sú konkrétne operácie násobenia v konvolučnách vrstvách. Experimentálne sme ukázali a porovnávali rozličné varianty a aj napriek tomu, že sme postupovali bez preučenia siete sa nám podarilo získať zaujímavé výsledky. Napríklad pri architektúre Inception v4 sme získali takmer 8% úspor, pričom nedošlo k žiadnemu poklesu presnosti. Táto úspora vie rozhodne nájsť uplatnenie v mobilných zariadeniach alebo pri veľkých neurónových sieťach s enormnými výpočtovými nárokmi.
|
|
|
Image Bit-Depth Expansion Method Based On Sparse
Representations
Záviška, Pavel
In this paper, a method for restoration of low bit-depth images based on sparse representations is presented. Proposed method is independent on the transform used, however for the purposes of experiments, Discrete Cosine Transform (DCT) and Discrete Wavelet Transform (DWT) are used. The experiments show that our method enhances the visual quality of low bit-depth images and performs better, in terms of PSNR, than basic bit-depth expansion methods.
|
|
|
Neural Network Implementation without Multiplication
Slouka, Lukáš ; Baskar, Murali Karthick (referee) ; Szőke, Igor (advisor)
The subject of this thesis is neural network acceleration with the goal of reducing the number of floating point multiplications. The theoretical part of the thesis surveys current trends and methods used in the field of neural network acceleration. However, the focus is on the binarization techniques which allow replacing multiplications with logical operators. The theoretical base is put into practice in two ways. First is the GPU implementation of crucial binary operators in the Tensorflow framework with a performance benchmark. Second is an application of these operators in simple image classifier. Results are certainly encouraging. Implemented operators achieve speed-up by a factor of 2.5 when compared to highly optimized cuBLAS operators. The last chapter compares accuracies achieved by binarized models and their full-precision counterparts on various architectures.
|
|
|
Modern compression formats of digital image
Eder, Radek ; Přinosil, Jiří (referee) ; Číka, Petr (advisor)
This bachelor thesis deals with the problems of digital image compression. There are described basic color models RGB and YCbCr and conversion between them. Described are currently used codecs for digital images. JPEG compression format is discussed in more detail, which currently is still the most widely used format for digital photos. Described is a process of digitizing the image using the JPEG standard, which includes sampling, quantization and coding. Described are also formats JPEG2000, PNG and WebP. End of chapter are described methods for subjective and objective evaluation of the quality of digital image compression. Among the subjective methods include Double Stimulus Impairment Scale (DSIS) and Double Stimulus Continuous Quality Scale (DSCQS), then the objective methods include Mean Squared Error (MSE), Peak Signal to Noise Ratio (PSNR) a Structural Similarity (SSIM). Application was created as part of the work that allows you to compress the source image to the selected formats. It also allows a qualitative evaluation of compression and store the results to the file.
|
|
|
Compression of ECG signal
Blaschová, Eliška ; Němcová, Andrea (referee) ; Vítek, Martin (advisor)
This paper represents the most well-known compression methods, which have been published. A Compression of ECG signal is important primarily for space saving in memory cards or efficiency improvement of data transfer. An application of wavelet transform for compression is a worldwide discussed topic and this is the reason why the paper focuses in this direction. Gained wavelet coefficients might be firstly quantized and then compressed using suitable method. There are many options for a selection of wavelet and a degree of decomposition, which will be tested from the point of view of the most efficient compression of ECG signal.
|
|
|
The USB Oscilloscope and Logic Analyzer with a Microprocessor
Pešl, Jiří ; Magát, Martin (referee) ; Šteffan, Pavel (advisor)
The content of this bachelor thesis is design of electronic measurement and diagnostic device to a computer or any other portable device as are modern mobile phones and PDAs. Designed system will be measure analog periodical signals in time in currently used ranges and discrete logic signals. Signals will be restored and displayed on the screen of any other device. Wherewith designed device will be communicate through USB 2.0 or Bluetooth interface.
|
guest ::
