Sökning: "Discriminative Latent Variable Models"

Hittade 3 avhandlingar innehållade orden Discriminative Latent Variable Models.

  1. 1. Visual Representations and Models: From Latent SVM to Deep Learning

    Författare :Hossein Azizpour; Stefan Carlsson; Barbara Caputo; KTH; []
    Nyckelord :TEKNIK OCH TEKNOLOGIER; ENGINEERING AND TECHNOLOGY; TEKNIK OCH TEKNOLOGIER; ENGINEERING AND TECHNOLOGY; Computer Vision; Machine Learning; Artificial Intelligence; Deep Learning; Learning Representation; Deformable Part Models; Discriminative Latent Variable Models; Convolutional Networks; Object Recognition; Object Detection; Computer Science; Datalogi;

    Sammanfattning : Two important components of a visual recognition system are representation and model. Both involves the selection and learning of the features that are indicative for recognition and discarding those features that are uninformative. LÄS MER

  3. 2. Improving Image Classification Performance using Joint Feature Selection

    Författare :Heydar Maboudi Afkham; Stefan Carlsson; Josef Kittler; KTH; []
    Nyckelord :NATURVETENSKAP; NATURAL SCIENCES; Image Classification; Latent Variable Models; Computer Science; Datalogi;

    Sammanfattning : In this thesis, we focus on the problem of image classification and investigate how its performance can be systematically improved. Improving the performance of different computer vision methods has been the subject of many studies. LÄS MER

  4. 3. Predicting Linguistic Structure with Incomplete and Cross-Lingual Supervision

    Författare :Oscar Täckström; Joakim Nivre; Jussi Karlgren; Ryan McDonald; Hal Daumé III; Uppsala universitet; []
    Nyckelord :NATURVETENSKAP; NATURAL SCIENCES; NATURVETENSKAP; NATURAL SCIENCES; linguistic structure prediction; structured prediction; latent-variable model; semi-supervised learning; multilingual learning; cross-lingual learning; indirect supervision; partial supervision; ambiguous supervision; part-of-speech tagging; dependency parsing; named-entity recognition; sentiment analysis; Computational Linguistics; Datorlingvistik;

    Sammanfattning : Contemporary approaches to natural language processing are predominantly based on statistical machine learning from large amounts of text, which has been manually annotated with the linguistic structure of interest. However, such complete supervision is currently only available for the world's major languages, in a limited number of domains and for a limited range of tasks. LÄS MER