Discriminative models for multi-class object layout
Chaitanya Desai, Deva Ramanan, Charless C. Fowlkes · IEEE International Conference on Computer Vision (ICCV) · 2009
Many state-of-the-art approaches for object recognition reduce the problem to a 0-1 classification task. Such reductions allow one to leverage sophisticated classifiers for learning. These models are typically trained independently for each class using positive and negative examples cropped from images. At test-time, various post-processing heuristics such as non-maxima suppression (NMS) are required to reconcile multiple detections within and between different classes for each image. Though crucial to good performance on benchmarks, this post-processing is usually defined heuristically. We introduce a unified model for multi-class object recognition that casts the problem as a structured prediction task. Rather than predicting a binary label for each image window independently, our model simultaneously predicts a structured labeling of the entire image. Our model learns statistics that capture the spatial arrangements of various object classes in real images, both in terms of which arrangements to suppress through NMS and which arrangements to favor through spatial co-occurrence statistics. We formulate parameter estimation in our model as a max-margin learning problem. Given training images with ground-truth object locations, we show how to formulate learning as a convex optimization problem. We employ a cutting plane algorithm similar to efficiently learn a model from thousands of training images. We show state-of-the-art results on the PASCAL VOC benchmark that indicate the benefits of learning a global model encapsulating the spatial layout of multiple object classes.
Cited by 274 works. doi:10.1109/iccv.2009.5459256
Authors
References
- Histograms of Oriented Gradients for Human Detection
- Robust Real-Time Face Detection
- Neural network-based face detection
- The PASCAL visual object classes challenge 2006 (VOC2006) results
- A discriminatively trained, multiscale, deformable part model
- TextonBoost: Joint Appearance, Shape and Context Modeling for Multi-class Object Recognition and Segmentation
- The Hidden Dimension
- Combined Object Categorization and Segmentation With An Implicit Shape Model
- An analysis of approximations for maximizing submodular set functions—I
- Multiscale conditional random fields for image labeling