Convolutional Kernel Networks for Action Recognition in Videos
Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
While convolutional neural networks (CNNs) have taken the lead for many learning tasks, action recognition in videos has yet to see this jump in performance. Many teams are working on the issue but so far there is no definitive answer how to make CNNs work well with video data. Recently, introduced convolutional kernel networks, a special case of CNNs which can be trained layer by layer in an unsupervised manner. This is done by approximating a kernel function in every layer with finite-dimensional descriptors. In this work we show the application of the CKN training to video, discuss the adjustments necessary and the influence of the type of data presented to the networks as well as the number of filters used.
Place, publisher, year, edition, pages
2015. , 39 p.
Convolutional Kernel Networks, Convolutional Neural Networks, Kernel Methods, Action Recognition, Computer Vision, Video
IdentifiersURN: urn:nbn:se:kth:diva-175797OAI: oai:DiVA.org:kth-175797DiVA: diva2:862363
Master of Science - Machine Learning
2015-09-29, Stockholm, 10:15 (English)
Sullivan, JosephineMairal, JulienSchmid, Cordelia