Sökning: "data augmentation"

Visar resultat 1 - 5 av 63 uppsatser innehållade orden data augmentation.

  1. 1. Deep Learning for Deep Water: Robust classification of ship wakes with expert in the loop

    Master-uppsats, Göteborgs universitet/Institutionen för data- och informationsteknik

    Författare :Igor RYAZANOV; [2020-10-06]
    Nyckelord :machine learning; deep learning; pattern recognition; acoustic data analysis; shipping data; data augmentation; noise robustness; classification with data imbalance; expert-in-the-loop framework;

    Sammanfattning : This work examines the applicability of the deep learning models to pattern recognitionin acoustic ocean data. The features of the dataset include noise, data scarcityand the lack of labeled samples. A deep learning model is proposed for the task ofautomatic wake detection. LÄS MER

  2. 2. Machine Learning for Detecting Hate Speech in Low Resource Languages

    Master-uppsats, Göteborgs universitet/Institutionen för data- och informationsteknik

    Författare :David Rodriguez; Denitsa Saynova; [2020-07-08]
    Nyckelord :machine learning; natural language processing; BERT; cross-lingual zeroshot learning; data augmentation; hate speech; classification; Twitter;

    Sammanfattning : This work examines the role of both cross-lingual zero-shot learning and data augmentationin detecting hate speech online for low resource set-ups. The proposedsolutions for situations where the amount of labeled data is scarce are to use alanguage with more resources during training or to create synthetic data points. LÄS MER

  3. 3. Building Detection in Deformed Satellite Images Using Mask R-CNN

    Master-uppsats, Blekinge Tekniska Högskola/Institutionen för datavetenskap

    Författare :Gayatri Chitturi; [2020]
    Nyckelord :Building detection; Mask R-CNN; Test Time Augmentation.;

    Sammanfattning : Background: In the recent research of automatic building detection, aerial and satellite images are used. Automatic building detection from satellite images is useful for urban planning, after natural disasters for identifying the voids. LÄS MER

  4. 4. Enhancing decision tree accuracy and compactness with improved categorical split and sampling techniques

    Master-uppsats, KTH/Skolan för elektroteknik och datavetenskap (EECS)

    Författare :Gaëtan Millerand; [2020]
    Nyckelord :;

    Sammanfattning : Decision tree is one of the most popular algorithms in the domain of explainable AI. From its structure, it is simple to induce a set of decision rules which are totally understandable for a human. That is why there is currently research on improving decision or mapping other models into a tree. Decision trees generated by C4. LÄS MER

  5. 5. Blood Cell Data Augmentation using Deep Learning Methods

    Master-uppsats, Lunds universitet/Matematik LTH

    Författare :Oskar Klang; Martin Carlberg; [2020]
    Nyckelord :Image analysis; deep learning; data augmentation; generative adversarial networks; GAN; Mathematics and Statistics;

    Sammanfattning : In this thesis we aim to improve classification performance on blood cell imagesby using deep learning techniques to augment data. The thesis was conductedat CellaVision, a company providing digital solutions for medical microscopy inthe field of hematology. LÄS MER