Essays about: "Latent Space"
Showing result 21 - 25 of 74 essays containing the words Latent Space.
-
21. Towards Latent Space Disentanglement of Variational AutoEncoders for Language
University essay from Uppsala universitet/Institutionen för lingvistik och filologiAbstract : Variational autoencoders (VAEs) are a neural network architecture broadly used in image generation (Doersch 2016). VAEs are neural network models that encode data from some domain and project it into a latent space (Doersch 2016). READ MORE
-
22. Computer Vision in Fitness: Exercise Recognition and Repetition Counting
University essay from KTH/Matematik (Avd.)Abstract : Motion classification and action localization have rapidly become essential tasks in computer vision and video analytics. In particular, Human Action Recognition (HAR), which has important applications in clinical assessments, activity monitoring, and sports performance evaluation, has drawn a lot of attention in research communities. READ MORE
-
23. Hybrid Variational Autoencoder for Clustering of Single-Cell RNA-seq Data : Introducing HybridVI, a Variational Autoencoder with two Latent Spaces
University essay from KTH/Skolan för elektroteknik och datavetenskap (EECS)Abstract : Single-cell analysis means to analyze cells on an individual level. This individual analysis enhances the investigation of the heterogeneity among and the classification of individual cells. Single-cell analysis is a broad term and can include various measurements. READ MORE
-
24. The cooling potential of Rain Gardens and Green Roofs - The influence on energy partitioning in Kvillebäcken, Gothenburg
University essay from Göteborgs universitet/Institutionen för geovetenskaperAbstract : Blue-green infrastructure (BGI) has become a popular way of combating the effects of climate change, both regarding heat mitigation and runoff from precipitation. This thesis aims to understand the potential of two BGI types, namely rain gardens and green roofs, of increasing the latent heat fraction in an urban area and thereby also understand the possible cooling effect from these measures. READ MORE
-
25. Segmenting the Left Atrium in Cardic CT Images using Deep Learning
University essay from Linköpings universitet/Institutionen för datavetenskapAbstract : Convolution neural networks have achieved a state of the art accuracy for multi-class segmentation in biomedical image science. In this thesis, a 2-Stage binary 2D UNet and MultiResUNet are used to segment the 3D cardiac CT Volumes. 3D volumes have been sliced into 2D images. READ MORE