Home > Search Results: Hajič, Jakub
Search:
 
Search Tips :: Simple Search
Search collections:
Sort by: Display results: Output format:
National Repository of Grey Literature 2 records found  Search took 0.00 seconds. 
Visual Question Answering
 Hajič, Jakub ; Straka, Milan (advisor) ; Lokoč, Jakub (referee)
Visual Question Answering (VQA) is a recently proposed multimodal task in the general area of machine learning. The input to this task consists of a single image and an associated natural language question, and the output is the answer to that question. In this thesis we propose two incremental modifications to an existing model which won the VQA Challenge in 2016 using multimodal compact bilinear pooling (MCB), a novel way of combining modalities. First, we added the language attention mechanism, and on top of that we introduce an image attention mechanism focusing on objects detected in the image ("region attention"). We also experiment with ways of combining these in a single end- to-end model. The thesis describes the MCB model and our extensions and their two different implementations, and evaluates them on the original VQA challenge dataset for direct comparison with the original work. 1
Detailed record
Extracting Control Points from Image Pairs for Perspective Transformation
 Hajič, Jakub ; Blažek, Jan (advisor) ; Soukup, Jindřich (referee)
Image registration is a part of many higher level image processing tasks. This thesis presents a novel approach to control point extraction as part of pairwise image registration. In combination with a suitable matching and model estimation algorithm, the extracted points could be used for estimating a global perspective transformation between the images. Our method extracts several features from image patches surrounding every point, and calculates an interest measure based on them. We use a supervised learning algorithm to obtain parameters of this extraction method. A control point matching algorithm is presented which considers the surrounding image patches of the control points and calculates a rotationally invariant similarity measure. We compare the results between our method of control point extraction and the Harris corner detector, and discuss the results of the matching and methods to improve both the matching and the control point extraction.
Detailed record

Interested in being notified about new results for this query?
Subscribe to the RSS feed.
Digital Repository :: Search :: Submit :: Personalize :: Help
Powered by Invenio v1.1.2
Maintained by admin.invenio@techlib.cz
This site is also available in the following languages:
Česky  English