Deva Ramanan
· University of California, Irvine
Is this your research profile?
Create or sign in to your KnowledgeTrend account to claim this page. After the claim, this same profile URL and its linked publications will belong to your account. Claiming does not automatically grant a verified badge.
Create account and claim this profile Sign in to claimMetrics are calculated from publications currently linked to this profile.
About
Deva Ramanan is a registered researcher in their academic field.
Recent Publications
2 research works linked to this profile
Object Detection with Discriminatively Trained Part-Based Models
We describe an object detection system based on mixtures of multiscale deformable part models. Our system is able to represent highly variable object classes and achieves state-of-the-art results in the PASCAL object detection challenges. While deformable part models have become quite popular, their value had not been demonstrated on difficult benchmarks such as the PASCAL data sets. Our system relies on new methods for discriminative training with partially labeled data. We combine a margin-sensitive approach for data-mining hard negative examples with a formalism we call latent SVM. A latent SVM is a reformulation of MI--SVM in terms of latent variables. A latent SVM is semiconvex, and the training problem becomes convex once latent information is specified for the positive examples. This leads to an iterative training algorithm that alternates between fixing latent values for positive examples and optimizing the latent SVM objective function.