0.25
0.5
0.75
1.25
1.5
1.75
2
Presentation
Deep Learning00:00
Currect Student and Postdocs13:25:30
Mining for Structure - 119:24:48
Mining for Structure - 233:53:25
Deep Generative Model36:48:57
Multimodal Data54:36:44
Example: Understanding Images58:55:13
Caption Generation86:25:14
Talk Roadmap - 194:44:36
Learning Feature Representations - 1107:50:15
Learning Feature Representations - 2113:04:02
Traditional Approaches117:57:25
Computer Vision Features - 1124:19:45
Computer Vision Features - 2131:34:17
Audio Features - 1135:51:29
Audio Features - 2136:40:53
Restricted Boltzmann Machines138:31:22
Learning features148:30:04
Model Learning - 1157:47:01
Model Learning - 2179:22:35
RBMs for Real-valued Data - 1190:23:49
RBMs for Real-valued Data - 2199:50:23
RBMs for Real-valued Data - 3204:03:43
RBMs for Word Counts - 1210:18:54
RBMs for Word Counts - 2228:00:32
Different Data Modalities240:15:22
Product of Experts - 1270:24:35
Product of Experts - 2307:41:51
Deep Boltzmann Machines - 1326:11:26
Deep Boltzmann Machines - 2330:07:09
Model Formulation341:00:02
Mathematical Formulation - 1373:48:32
Mathematical Formulation - 2380:16:38
Mathematical Formulation - 3390:27:40
Approximate Learning - 1408:14:25
Approximate Learning - 411:25:53
Approximate Learning - 3438:05:42
Previous Work442:52:32
New Learning Algorithm - 1460:00:47
New Learning Algorithm - 2471:02:51
New Learning Algorithm - 3509:39:09
Stochastic Approximation518:25:21
Learning Algorithm550:44:45
Variational Inference - 1589:27:20
Variational Inference - 2668:13:28
Variational Inference - 3685:06:28
Variational Inference - 4693:10:17
Good Generative Model? - 1695:03:27
Good Generative Model? - 2705:43:08
Good Generative Model? - 3711:14:32
Good Generative Model? - 4713:33:05
Good Generative Model? - 5722:45:13
Good Generative Model? - 6731:19:22
Handwriting Recognition743:53:47
Generative Model of 3-D Objects753:20:29
3-D Object Recognition765:25:54
Talk Roadmap - 2778:07:44
Data - Collection of Modalities782:36:05
Shared Concept790:05:23
Multi-Modal Input792:57:03
Challenges - I801:15:32
Challenges - II - 1812:18:37
Challenges - II - 2818:29:55
A Simple Multimodal Model821:23:09
Multimodal DBM - 1828:09:46
Multimodal DBM - 2829:07:50
Multimodal DBM - 3829:14:49
Multimodal DBM - 4832:55:36
Multimodal DBM - 5843:06:01
Text Generated from Images - 1853:06:13
Text Generated from Images - 2871:58:52
Images from Text883:34:08
MIR-Flickr Dataset889:12:37
Data and Architecture893:47:28
Results - 1906:43:49
Results - 2910:15:06
Generating Sentences915:18:01
Talk Roadmap - 3945:54:34
Markov Random Fields958:08:44
Restricted Boltzmann Machines974:12:14
Model Selection988:38:25
Generative Model1001:30:16
Model Selection - 11017:28:30
Model Selection - 21021:49:56
Simple Importance Sampling1027:47:08
Annealing Between Distributions - 11088:48:02
Annealing Between Distributions - 21134:11:45
Annealed Importance Sampling Run1143:37:15
AIS is Importance Sampling - 11159:14:37
AIS is Importance Sampling - 21240:13:18
RBMs with Geometric Averages1244:15:08
Problems with Undirected Models1254:32:21
Motivation: RBM Sampling - 11291:56:49
Motivation: RBM Sampling - 21293:08:14
Motivation: RBM Sampling - 31294:30:34
Motivation: RBM Sampling - 41294:47:01
Motivation: RBM Sampling - 51295:01:55
Motivation: RBM Sampling - 61295:16:02
Motivation: RBM Sampling - 71297:46:07
Unrolled RBM as a Deep Generative Model - 11302:02:35
Unrolled RBM as a Deep Generative Model - 21302:52:07
Unrolled RBM as a Deep Generative Model - 31302:57:55
Unrolled RBM as a Deep Generative Model - 41303:06:49
Unrolled RBM as a Deep Generative Model - 51303:13:12
Unrolled RBM as a Deep Generative Model - 61305:36:35
Reverse AIS Estimator (RAISE) - 11308:34:45
Reverse AIS Estimator (RAISE) - 21317:32:58
Reverse AIS Estimator (RAISE) - 31328:14:26
MNIST - 11340:12:10
MNIST - 21352:15:03
Omniglot Dataset1356:25:08
MNIST and Omniglot Results1360:52:00
DBMs and DBNs1369:28:24
Helmholtz Machines1377:39:41