|
Current Projects
Semantic Concept Detection with Cross Domain Learning
Introduction
Exploding amounts of multimedia data increasingly require automatic indexing and classification,
e.g. training classifiers to produce high-level features, or semantic concepts, chosen to represent image
content, like car, person, etc. When changing the applied domain (i.e. from news domain to consumer home
videos), the classifiers trained in one domain often perform poorly in the other domain due to changes
in feature distributions. Additionally, classifiers trained on the new domain alone may suffer from
too few positive training samples. Appropriately adapting data/models from an old domain to help
classify data in a new domain is an important issue.
Cross Domain SVM
We develop a new cross-domain SVM (CDSVM) algorithm for adapting previously learned support vectors
from one domain to help classification in another domain. Better precision is obtained with almost
no additional computational cost. Also, we give a comprehensive summary and comparative study of the
state-of-the-art SVM-based cross-domain learning methods. Evaluation over the latest large-scale
TRECVID benchmark data set shows that our CDSVM method can improve mean average precision over 36
concepts by 7.5%. For further performance gain, we also propose an intuitive selection criterion
to determine which cross-domain learning method to use for each concept.
|
|
Semantic Event Detection based on Concept Prediction
We investigate automatic detection of semantic events in users' image and video collections.
A novel semantic event detection approach is proposed which exploits an event-level
Bag-of-Features (BOF) representation to model typical events. Based on this BOF representation,
semantic events are detected in a concept space instead of the original
low-level visual feature space. There are three advantages of our approach: we can avoid the
sensitivity problem by decreasing the influence of difficult or erroneous images or videos in
measuring the event-level similarity; also we can utilize the power of higher-level concept scores
in describing semantic events; in addition all different types of concept detectors trained from different
previous domains can be incorporated to generate the concept score for helping detection in the current
event data set, and these concept prediction scores play the role of transferring information learned in previous
domain to the current domain. Preliminary experiments are conducted to use concept detectors trained from Kodak's videos
to help detect semantic events in the consumer event collection.
|
Current Work
I am current developing a novel algorithm to explore cross-domain learning and semi-supervised learning together and to exert
advantages of both techniques in multimedia concept detection.
|
Publications
- Wei Jiang, Eric Zavesky, Shih-Fu Chang, Alexander C. Loui, "Cross-domain learning methods for
high-level visual concept classification", ICIP, San Diego, CA, 2008.
PDF
- Wei Jiang, Alexander C. Loui, "Semantic event detection based on visual concept prediction", ICME,
Germany, 2008. PDF
Patent
- Alexander C. Loui, Wei Jiang, Semantic Event Detection for Digital Content Records, US Patent
Pending, Filed Dec. 2008.
|