Stefanie Jegelka

Deep Metric Learning via Lifted Structured Feature Embedding

Learning the distance metric between pairs of examples is of great importance for learning and visual recognition. With the remarkable success from the state of the art convolutional neural networks, recent works [1, 31] have shown promising results on discriminatively training the networks to learn semantic feature embeddings where similar examples are mapped close to each other and dissimilar examples are mapped farther apart. In this paper, we describe an algorithm for taking full advantage of the training batches in the neural network training by lifting the vector of pairwise distances within the batch to the matrix of pairwise distances. This step enables the algorithm to learn the state of the art feature embedding by optimizing a novel structured prediction objective on the lifted problem. Additionally, we collected Stanford Online Products dataset: 120k images of 23k classes of online products for metric learning. Our experiments on the CUB-200-2011 [37], CARS196 [19], and Stanford Online Products datasets demonstrate significant improvement over existing deep feature embedding methods on all experimented embedding sizes with the GoogLeNet [33] network. The source code and the dataset are available at: https://github.com/rksltnl/ Deep-Metric-Learning-CVPR16.

Submodularity beyond submodular energies: Coupling edges in graph cuts

We propose a new family of non-submodular global energy functions that still use submodularity internally to couple edges in a graph cut. We show it is possible to develop an efficient approximation algorithm that, thanks to the internal submodularity, can use standard graph cuts as a subroutine. We demonstrate the advantages of edge coupling in a natural setting, namely image segmentation. In particular, for fine-structured objects and objects with shading variation, our structured edge coupling leads to significant improvements over standard approaches.

Fast Kernel-Based Independent Component Analysis

Recent approaches to independent component analysis (ICA) have used kernel independence measures to obtain highly accurate solutions, particularly where classical methods experience difficulty (for instance, sources with near-zero kurtosis). FastKICA (fast HSIC-based kernel ICA) is a new optimization method for one such kernel independence measure, the Hilbert-Schmidt Independence Criterion (HSIC). The high computational efficiency of this approach is achieved by combining geometric optimization techniques, specifically an approximate Newton-like method on the orthogonal group, with accurate estimates of the gradient and Hessian based on an incomplete Cholesky decomposition. In contrast to other efficient kernel-based ICA algorithms, FastKICA is applicable to any twice differentiable kernel function. Experimental results for problems with large numbers of sources and observations indicate that FastKICA provides more accurate solutions at a given cost than gradient descent on HSIC. Comparing with other recently published ICA methods, FastKICA is competitive in terms of accuracy, relatively insensitive to local minima when initialized far from independence, and more robust towards outliers. An analysis of the local convergence properties of FastKICA is provided.

A Principled Deep Random Field Model for Image Segmentation

We discuss a model for image segmentation that is able to overcome the short-boundary bias observed in standard pairwise random field based approaches. To wit, we show that a random field with multi-layered hidden units can encode boundary preserving higher order potentials such as the ones used in the cooperative cuts model of [11] while still allowing for fast and exact MAP inference. Exact inference allows our model to outperform previous image segmentation methods, and to see the true effect of coupling graph edges. Finally, our model can be easily extended to handle segmentation instances with multiple labels, for which it yields promising results.

Deep Metric Learning via Facility Location

Learning image similarity metrics in an end-to-end fashion with deep networks has demonstrated excellent results on tasks such as clustering and retrieval. However, current methods, all focus on a very local view of the data. In this paper, we propose a new metric learning scheme, based on structured prediction, that is aware of the global structure of the embedding space, and which is designed to optimize a clustering quality metric (NMI). We show state of the art performance on standard datasets, such as CUB200-2011 [37], Cars196 [18], and Stanford online products [30] on NMI and R@K evaluation metrics.

Solution stability in linear programming relaxations: graph partitioning and unsupervised learning

We propose a new method to quantify the solution stability of a large class of combinatorial optimization problems arising in machine learning. As practical example we apply the method to correlation clustering, clustering aggregation, modularity clustering, and relative performance significance clustering. Our method is extensively motivated by the idea of linear programming relaxations. We prove that when a relaxation is used to solve the original clustering problem, then the solution stability calculated by our method is conservative, that is, it never overestimates the solution stability of the true, unrelaxed problem. We also demonstrate how our method can be used to compute the entire path of optimal solutions as the optimization problem is increasingly perturbed. Experimentally, our method is shown to perform well on a number of benchmark problems.

Learning Scalable Discriminative Dictionary with Sample Relatedness

Attributes are widely used as mid-level descriptors of object properties in object recognition and retrieval. Mostly, such attributes are manually pre-defined based on domain knowledge, and their number is fixed. However, pre-defined attributes may fail to adapt to the properties of the data at hand, may not necessarily be discriminative, and/or may not generalize well. In this work, we propose a dictionary learning framework that flexibly adapts to the complexity of the given data set and reliably discovers the inherent discriminative middle-level binary features in the data. We use sample relatedness information to improve the generalization of the learned dictionary. We demonstrate that our framework is applicable to both object recognition and complex image retrieval tasks even with few training examples. Moreover, the learned dictionary also help classify novel object categories. Experimental results on the Animals with Attributes, ILSVRC2010 and PASCAL VOC2007 datasets indicate that using relatedness information leads to significant performance gains over established baselines.

Generalized clustering via kernel embeddings

We generalize traditional goals of clustering towards distinguishing components in a non-parametric mixture model. The clusters are not necessarily based on point locations, but on higher order criteria. This framework can be implemented by embedding probability distributions in a Hilbert space. The corresponding clustering objective is very general and relates to a range of common clustering concepts.

Prenatal development of ocular dominance and orientation maps in a self-organizing model of V1

How orientation and ocular-dominance (OD) maps develop before visual experience begins is controversial. Possible influences include molecular signals and spontaneous activity, but their contributions remain unclear. This paper presents LISSOM simulations suggesting that previsual spontaneous activity alone is sufficient for realistic OR and OD maps to develop. Individual maps develop robustly with various previsual patterns, and are aided by background noise. However, joint OR/OD maps depend crucially on how correlated the patterns are between eyes, even over brief initial periods. Therefore, future biological experiments should account for multiple activity sources, and should measure map interactions rather than maps of single features.

Efficient and accurate clustering for large-scale genetic mapping

High-throughput “next generation” genome sequencing technologies are producing a flood of inexpensive genetic information that is invaluable to genomics research. Sequences of millions of genetic markers are being produced, providing genomics researchers with the opportunity to construct highresolution genetic maps for many complicated genomes. However, the current generation of genetic mapping tools were designed for the small data setting, and are now limited by the prohibitively slow clustering algorithms they employ in the genetic marker-clustering stage. In this work, we present a new approach to genetic mapping based on a fast clustering algorithm that exploits the geometry of the data. Our theoretical and empirical analysis shows that the algorithm can correctly recover linkage groups. Using synthetic and real-world data, including the grand-challenge wheat genome, we demonstrate that our approach can quickly process orders of magnitude more genetic markers than existing tools while retaining - and in some cases even improving - the quality of genetic marker clusters.