This paper presents a novel physics-based representation of realistic character motion. The dynamical model incorporates several factors of locomotion derived from the biomechanical literature, including relative preferences for using some muscles more than others, elastic mechanisms at joints due to the mechanical properties of tendons, ligaments, and muscles, and variable stiffness at joints depending on the task. When used in a spacetime optimization framework, the parameters of this model define a wide range of styles of natural human movement.

Abstract not found

Abstract not found

We describe a novel method for simultaneously detecting faces and estimating their pose in real time. The method employs a convolutional network to map images of faces to points on a lowdimensional manifold parametrized by pose, and images of non-faces to points far away from that manifold. Given an image, detecting a face and estimating its pose is viewed as minimizing an energy function with respect to the face/non-face binary variable and the continuous pose parameters. The system is trained to minimize a loss function that drives correct combinations of labels and pose to be associated with lower energy values than incorrect ones. The system is designed to handle very large range of poses without retraining. The performance of the system was tested on three standard data sets—for frontal views, rotated faces, and profiles— is comparable to previous systems that are designed to handle a single one of these data sets. We show that a system trained simuiltaneously for detection and pose estimation is more accurate on both tasks than similar systems trained for each task separately. 1

Energy-Based Models (EBMs) capture dependencies between variables by associating a scalar energy to each configuration of the variables. Inference consists in clamping the value of observed variables and finding configurations of the remaining variables that minimize the energy. Learning consists in finding an energy function in which observed configurations of the variables are given lower energies than unobserved ones. The EBM approach provides a common theoretical framework for many learning models, including traditional discriminative and generative approaches, as well as graph-transformer networks, conditional random fields, maximum margin Markov networks, and several manifold learning methods. Probabilistic models must be properly normalized, which sometimes requires evaluating intractable integrals over the space of all possible variable configurations. Since EBMs have no requirement for proper normalization, this problem is naturally circumvented. EBMs can be viewed as a form of non-probabilistic factor graphs, and they provide considerably more flexibility in the design of architectures and training criteria than probabilistic approaches. 1

Markov Random Field (MRF) models are a popular tool for vision and image processing. Gaussian MRF models are particularly convenient to work with because they can be implemented using matrix and linear algebra routines. However, recent research has focused on on discrete-valued and non-convex MRF models because Gaussian models tend to over-smooth images and blur edges. In this paper, we show how to train a Gaussian Conditional Random Field (GCRF) model that overcomes this weakness and can outperform the non-convex Field of Experts model on the task of denoising images. A key advantage of the GCRF model is that the parameters of the model can be optimized efficiently on relatively large images. The competitive performance of the GCRF model and the ease of optimizing its parameters make the GCRF model an attractive option for vision and image processing applications. 1.

Abstract. Estimation of parameters of random field models from labeled training data is crucial for their good performance in many image analysis applications. In this paper, we present an approach for approximate maximum likelihood parameter learning in discriminative field models, which is based on approximating true expectations with simple piecewise constant functions constructed using inference techniques. Gradient ascent with these updates exhibits compelling limit cycle behavior which is tied closely to the number of errors made during inference. The performance of various approximations was evaluated with different inference techniques showing that the learned parameters lead to good classification performance so long as the method used for approximating the gradient is consistent with the inference mechanism. The proposed approach is general enough to be used for the training of, e.g., smoothing parameters of conventional Markov Random Fields (MRFs). 1

Side-chain prediction is an important subproblem of the general protein folding problem. Despite much progress in side-chain prediction, performance is far from satisfactory. As an example, the ROSETTA protocol that uses simulated annealing to select the minimum energy conformations, correctly predicts the first two side-chain angles for approximately 72 % of the buried residues in a standard data set. Is further improvement more likely to come from better search methods, or from better energy functions? Given that exact minimization of the energy is NP hard, it is difficult to get a systematic answer to this question. In this paper, we present a novel search method and a novel method for learning energy functions from training data that are both based on Tree Reweighted Belief Propagation (TRBP). We find that TRBP can find the global optimum of the ROSETTA energy function in a few minutes of computation for approximately 85 % of the proteins in a standard benchmark set. TRBP can also effectively bound the partition function which enables using the Conditional Random Fields (CRF) framework for learning. Interestingly, finding the global minimum does not significantly improve side-chain prediction for

submitted for publication Abstract. Text categorization and retrieval tasks are often based on a good representation of textual data. Departing from the classical vector space model, several probabilistic models have been proposed recently, such as PLSA and LDA. In this paper, we propose the use of a neural network based, non-probabilistic, solution, which captures jointly a rich representation of words and documents. Experiments performed on two information retrieval tasks using the TDT2 database and the TREC-8 and 9 sets of queries yielded a better performance for the proposed neural network model, as compared to PLSA and the classical TFIDF representations. 2 IDIAP–RR 05-12

Spectral clustering and eigenvector-based methods have become increasingly popular in segmentation and recognition. Although the choice of the pairwise similarity metric (or affinities) greatly influences the quality of the results, this choice is typically specified outside the learning framework. In this paper, we present an algorithm to learn class-specific similarity functions. Mapping our problem in a Conditional Random Fields (CRF) framework enables us to pose the task of learning affinities as parameter learning in undirected graphical models. There are two significant advances over previous work. First, we learn the affinity between a pair of data-points as a function of a pairwise feature and (in contrast with previous approaches) the classes to which these two data-points were mapped, allowing us to work with a richer class of affinities. Second, our formulation provides a principled probabilistic interpretation for learning all of the parameters that define these affinities. Using ground truth segmentations and labellings for training, we learn the parameters with the greatest discriminative power (in an MLE sense) on the training data. We demonstrate the power of this learning algorithm in the setting of joint segmentation and recognition of object classes. Specifically, even with very simple appearance features, the proposed method achieves state-of-the-art performance on standard datasets. 1.