|
|
Codec Detection from Speech
Jon, Josef ; Matějka, Pavel (referee) ; Černocký, Jan (advisor)
Tato práce se zabývá detekcí kodeků z komprimovaného řečového signálu. Cílem bylo zjistit, jaké charakteristiky rozlišují jednotlivé kodeky a následně vytvořit prostředí vhodné pro experimenty s různými typy a konfiguracemi klasifikátorů. Použity byly Support vector machines a především neuronové sítě, které byly vytvořeny pomocí nástroje Keras. Hlavním přínosem této práce je experimentální část, ve které je analyzován vliv různých parametrů neuronové sítě. Po nalezení nejvhodnější kombinace parametrů dosáhla síť přesnosti klasifikace přes 98% na testovací sadě obsahující data z 6 kodeků.
|
|
|
Deep Neural Networks for Classifying Objects in an Image
Mlynarič, Tomáš ; Zemčík, Pavel (referee) ; Hradiš, Michal (advisor)
This paper deals with classifying objects using deep neural networks. Whole scene segmentation was used as main algorithm for the classification purpose which works with video sequences and obtains information between two video frames. Optical flow was used for getting information from the video frames, based on which features maps of a~neural network are warped. Two neural network architectures were adjusted to work with videos and experimented with. Results of the experiments show, that using videos for image segmentation improves accuracy (IoU) compared to the same architecture working with images.
|
|
|
Fine-Grained Vehicle Recognition from Traffic Surveillance Camera
Mencner, Pavel ; Špaňhel, Jakub (referee) ; Sochor, Jakub (advisor)
The aim of this thesis is image based detection of vehicles from traffic surveillance camera and fine-grained vehicle type recognition (manufacturer and model). In the thesis the Unpack normalization method is implemented which transforms the vehicle image into its apparent flat representation in order to increase the classifier's success rate. The Unpack method make use of 3D bounding box of the vehicle. This bounding box is constructed during test period using the information of vehicle contour and direction toward vanishing points. The thesis involve accuracy comparison between direct and Unpack classification methods. The proposed solution is based on several related parts that benefit from convolutional neural networks. These parts are: vehicle detection from image data, estimation of the directions towards vanishing points solved as classification task, vehicle contour detection using convolutional Encoder-Decoder network and fine-grained vehicle type classification. Using Unpack based classification the 2% accuracy improvement against direct classification has been achieved, resulting in 86% overall success rate. The outcome of this thesis is fine-grained vehicle classification system that works with traffic surveillance video without any viewpoint limitations.
|
|
|
Smartphone Game Using Recognition of Face Features
Skoták, Jiří ; Szőke, Igor (referee) ; Herout, Adam (advisor)
This master's thesis focuses on smartphone game for iOS, which uses recognition of face features and other information, which can be obtained from a smartphone's camera and sensors. This work describes a few approaches for real-time face detection and then introduces and compares possibilities for such task on iOS. Moreover, the thesis contains a draft of the final game and its levels. The game showcases various technologies in its levels such as object detection, processing an image color and others. Finally, the thesis introduces the final form of the game that is released and available on the App Store.
|
|
|
Deep learning methods for vessel and optic disc segmentation in ophthalmologic sequences
Rozhoňová, Andrea ; Odstrčilík, Jan (referee) ; Hesko, Branislav (advisor)
The aim of the following thesis was to study the issue of optical disc and retinal vessels segmentation in ophthalmologic sequences. The theoretical part of the thesis summarizes the principles of different approaches in the field of deep learning, which are used in connection with the given issue. Based on the theoretical part, methods for optical disk segmentation and retinal vessel segmentation based on the convolutional neural networks Linknet, PSPNet, Unet and MaskRCNN are proposed. The practical part of the thesis deals with the description of their implementation and subsequent evaluation.
|
|
|
Audio noise reduction using deep neural networks
Talár, Ondřej ; Galáž, Zoltán (referee) ; Harár, Pavol (advisor)
The thesis focuses on the use of deep recurrent neural network, architecture Long Short-Term Memory for robust denoising of audio signal. LSTM is currently very attractive due to its characteristics to remember previous weights, or edit them not only according to the used algorithms, but also by examining changes in neighboring cells. The work describes the selection of the initial dataset and used noise along with the creation of optimal test data. For creation of the training network is selected KERAS framework for Python and are explored and discussed possible candidates for viable solutions.
|
|
|
Face Detection in Poor Quality Videos
Koval, Michal ; Orság, Filip (referee) ; Goldmann, Tomáš (advisor)
This bachelor thesis deals with face detection in low quality videos, while mainly focusing on occluded faces. It describes elementary priciples of machine learning algorithms and their methods, which are often used in the field of computer vision. Out of them are more closely described convolutional neural networks and their state of the art models focused on face detection. Out of those, convolutional neural networks and state of the art models for face detection are more closely described. For the practical part face detection models inspired by state of the art model RetinaFace were implemented and trained. The best performing model achieves 85.5% average precision on WIDER Face HARD testing dataset and 90.9% on dataset focused on occluded faces. Part of this thesis is also a program with graphical user interfaces which provides tools to use developed models on videos and pictures.
|
|
|
Automated Identification of Graphical UI Elements for Robotic Quality Assurance
Válek, Lukáš ; Herout, Adam (referee) ; Španěl, Michal (advisor)
Tato práce se zabývá problematikou analýzy obrazovek grafického uživatelského rozhraní (GUI) pomocí konvolučních neuronových sítí (CNN) a metod počítačového vidění. Cílem této práce je vytvořit systém, který automaticky identifikuje GUI elementy na základě piktogramových a textových informací pro detekované prvky ve vstupním obrázku. K~vývoji systému byla použita kombinace EfficientNetB1 CNN, OCR a tradičních metod počítačového vidění. K trénování CNN byla použita vlastní datová sada, která obsahovala 120 tisíc piktogramů. Byl vytvořen sémantický slovník UI prvků, který dále využívá text detekovaný pomocí OCR. Nakonec byl vytvořen podsystém pro analýzu GUI hierarchie, který slouží k detekci a sémantické kategorizaci oblastí GUI. Výsledný systém automaticky klasifikuje detekované piktogramy, navrhuje další třídy na základě textu a rozděluje GUI obrazovku do hierarchických sekcí. Systém dosahuje 81,1% přesnosti identifikace UI prvků a v průměru zanalyzuje jednu obrazovku za 0,6 sekundy. Systém automatizuje identifikaci UI prvků, čímž umožňuje zaměstnancům věnovat se jiným činnostem. V budoucnu lze tento systém dále rozvíjet, aby sloužil jako základ pro automatické exploratorní testování.
|
|
|
Crowd Density Estimation from a Photo
Ferencz, Adam ; Herout, Adam (referee) ; Beran, Vítězslav (advisor)
The aim of this thesis is to develop an aplication estimating the total number of people at a demonstration or at other public events. Input is a serie of photos from a drone or some other photos. The output are couloured maps according to people density in the place. Photos are placed in a topological map. Convolutional neural network MCNN is used for the crowd counting, which can generate a density map from the photo. Special method was proposed to correct the total amount of counted people when photographs overlap. The application is divided into server and web client. The server part generates density maps, saves data and runs an overlap correction algorithm. Client handles user inputs and provides an interactiv map with visualization.
|
|
|
Protection of sensitive data contained in images
Mezina, Anzhelika ; Rajnoha, Martin (referee) ; Burget, Radim (advisor)
Tato bakalářská práce je zaměřena na využití hlubokého učení v bezpečnostním problému úniku citlivých informací ve formě obrazových dat. Pokusem o vyřešení tohoto problému bylo použití Single Shot Multibox Detectoru (SSD) a plně propojené sítě, poslední je mnohem rychlejší než jiné metody a může být použitá v praxi, kde je potřeba velmi rychlé analýzy příchozí a odchozí informace, například analýzy provozu sítě. V první části práce jsou popsané metody, které mohou být použité pro detekci klíčových slov. Druhá část obsahuje popis experimentu a dosažených výsledků pro dva modely neuronových sítí: Single Shot Multibox Detector a plně propojené sítě. Druhý model dosahuje uspokojivých vlastností jak z pohledu času zpracování tak i přesnosti a lze jej použít v praxi.
|
guest ::
