|
|
Procedural Generation and Simulation of 2D Gaming World
Dubský, Tomáš ; Kocur, Viktor (referee) ; Chlubna, Tomáš (advisor)
Cílem práce je implementace procedurálního generování a simulace dvojdimenzionálního herního světa. Herní svět je tvořen nekonečnou mřížkou malých dlaždic. Tyto dlaždice jsou seskupeny do částí, takže svět je generován a simulován pouze pro ty části, které jsou poblíž hráče. Generovaný terén se skládá z několika biomů a podzemních jeskyní. Kapaliny, plyny nebo třeba růst trávy patří mezi procesy, které jsou simulovány.
|
|
|
Self-Supervised Learning for Recognition of Sports Poses in Image
Olekšák, Samuel ; Kocur, Viktor (referee) ; Herout, Adam (advisor)
This thesis demonstrates a solution for minimizing the amount of necessary labelled training data in the classification of sports poses using a neural network trained with contrastive self-supervised learning. Training consists of two stages. The first stage trains a feature extractor which uses unlabelled training images extracted from recordings of exercises from multiple viewpoints. In the second stage, using a small amount of labelled data, a simple classifier connected to the feature extractor is trained. The thesis discusses classification in the context of yoga poses, however, the final solution can be easily applied to any other sport in case of obtaining a suitable dataset. During the development of the solution, emphasis is placed on the performance of the resulting model so that it can be used on mobile devices. The resulting model reached an accuracy of 76 % using augmentations with a data set containing four labelled images per yoga pose. On a larger data set with 800 labelled images for all poses, an accuracy of 82 % is reached.
|
|
|
Experiments with Estimation of Human Pose in Image and Video
Horejš, Michal ; Kocur, Viktor (referee) ; Herout, Adam (advisor)
The detection of phenomena in the image has a wide application in many fields and it is therefore important to constantly develop and improve the detection. This work specifically deals with the problem of detecting sports positions in images and videos. The goal was to experiment with tools for human pose recognition and general phenomenon recognition. During the experiments, three new data sets were created, in which key points of the human body are used. The datasets were then trained on several modeled architectures of convolutional neural networks. The results of the experiments show that the appropriate use of key points can help with the detection of sports positions.
|
|
|
Self-Supervised Learning for Recognition of Hand Poses in Image
Makaiová, Lucia ; Kocur, Viktor (referee) ; Herout, Adam (advisor)
This work focuses on using self-supervised learning for the task of hand poses recognition in image. I have used contrastive method of self-supervised learning and optimized the solution iteratively, using techniques such as early stopping, triplet mining, optimization of hyperparameters or experimenting with various model architectures. The method was implemented with Pytorch framework and Tensorboard was used for data processing and visualization. I have trained the first model using a supervised method, to obtain reference values. I have successfully matched this reference result by training a self-supervised model on Handz dataset and achieving 83% accuracy. The created solution provides findings, which can be applied to similar problems, such as recognition of sport poses. The main contribution of this work is the discovery, that self-supervised methods are particularly effective when using a labeled dataset for downstream task with just a small amount of samples, which in addition have uneven distribution of samples for individual classes. Based on these findings, it is possible to create a method for self-supervised learning for recognition of sport poses or further optimize existing solution for hand poses.
|
|
|
Self-Supervised Learning for Recognition of Sports Poses in Image
Olekšák, Samuel ; Kocur, Viktor (referee) ; Herout, Adam (advisor)
This thesis demonstrates a solution for minimizing the amount of necessary labelled training data in the classification of sports poses using a neural network trained with contrastive self-supervised learning. Training consists of two stages. The first stage trains a feature extractor which uses unlabelled training images extracted from recordings of exercises from multiple viewpoints. In the second stage, using a small amount of labelled data, a simple classifier connected to the feature extractor is trained. The thesis discusses classification in the context of yoga poses, however, the final solution can be easily applied to any other sport in case of obtaining a suitable dataset. During the development of the solution, emphasis is placed on the performance of the resulting model so that it can be used on mobile devices. The resulting model reached an accuracy of 76 % using augmentations with a data set containing four labelled images per yoga pose. On a larger data set with 800 labelled images for all poses, an accuracy of 82 % is reached.
|
|
|
Procedural Generation and Simulation of 2D Gaming World
Dubský, Tomáš ; Kocur, Viktor (referee) ; Chlubna, Tomáš (advisor)
Cílem práce je implementace procedurálního generování a simulace dvojdimenzionálního herního světa. Herní svět je tvořen nekonečnou mřížkou malých dlaždic. Tyto dlaždice jsou seskupeny do částí, takže svět je generován a simulován pouze pro ty části, které jsou poblíž hráče. Generovaný terén se skládá z několika biomů a podzemních jeskyní. Kapaliny, plyny nebo třeba růst trávy patří mezi procesy, které jsou simulovány.
|
guest ::
RSS feed.