This paper presents a robust multi-view method for tracking people in 3D scene. Our method distinguishes itself from previous works in two aspects. Firstly, we define a set of binary spatial relationships for individual subjects or pairs of subjects that appear at the same time, e.g. being left or right, being closer or further to the camera, etc. These binary relationships directly reflect relative positions of subjects in 3D scene and thus should be persisted during inference. Secondly, we introduce an unified probabilistic framework to exploit binary spatial constraints for simultaneous 3D localization and cross-view human tracking. We develop a cluster Markov Chain Monte Carlo method to search the optimal solution. We evaluate our method on both public video benchmarks and newly built multi-view video dataset. Results with comparisons showed that our method could achieve state-of-the-art tracking results and meter-level 3D localization on challenging videos.

We consider the problem of subspace clustering using the SSC (Sparse Subspace Clustering) approach, which has several desirable theoretical properties and has been shown to be effective in various computer vision applications.We develop a large scale distributed framework for the computation of SSC via an alternating direction method of multiplier (ADMM) algorithm. The proposed framework solves SSC in column blocks and only involves parallel multivariate Lasso regression subproblems and sample-wise operations. This appealing property allows us to allocate multiple cores/machines for the processing of individual column blocks.We evaluate our algorithm on a shared-memory architecture. Experimental results on real-world datasets confirm that the proposed block-wise ADMM framework is substantially more efficient than its matrix counterpart used by SSC,without sacrificing accuracy. Moreover, our approach is directly applicable to decentralized neighborhood selection for Gaussian graphical models structure estimation.

Feature selection is essential for effective visual recognition. We propose an efficient joint classifier learning and feature selection method that discovers sparse, compact representations of input features from a vast sea of candidates, with an almost unsupervised formulation. Our method requires only the following knowledge, which we call the feature sign - whether or not a particular feature has on average stronger values over positive samples than over negatives. We show how this can be estimated using as few as a single labeled training sample per class. Then, using these feature signs, we extend an initial supervised learning problem into an (almost) unsupervised clustering formulation that can incorporate new data without requiring ground truth labels. Our method works both as a feature selection mechanism and as a fully competitive classifier. It has important properties, low computational cost annd excellent accuracy, especially in difficult cases of very limited training data. We experiment on large-scale recognition in video and show superior speed and performance to established feature selection approaches such as AdaBoost, Lasso, greedy forward-backward selection, and powerful classifiers such as SVM.

Modeling complex crowd behaviour for tasks such as rare event detection has received increasing interest. However, existing methods are limited because (1) they are sensitive to noise often resulting in a large number of false alarms; and (2) they rely on elaborate models leading to high computational cost thus unsuitable for processing a large number of video inputs in real-time. In this paper, we overcome these limitations by introducing a novel complex behaviour modeling framework, which consists of a Binarized Cumulative Directional (BCD) feature as representation, novel spatial and temporal context modeling via an iterative correlation maximization, and a set of behaviour models, each being a simple Bernoulli distribution. Despite its simplicity, our experiments on three benchmark datasets show that it significantly outperforms the state-of-the-art for both temporal video segmentation and rare event detection. Importantly, it is extremely efficient — reaches 90Hz on a normal PC platform using MATLAB.

A key challenge in the realization of autonomous vehicles is the machine's ability to perceive its surrounding environment. This task is tackled through a model that partitions vehicle camera input into distinct semantic classes, by taking into account visual contextual cues. The use of structured machine learning models is investigated, which not only allow for complex input, but also arbitrarily structured output. Towards this goal, an outdoor road scene dataset is constructed with accompanying fine-grained image labelings. For coherent segmentation, a structured predictor is modeled to encode label distributions conditioned on the input images. After optimizing this model through max-margin learning, based on an ontological loss function, efficient classification is realized via graph cuts inference using alpha-expansion. Both quantitative and qualitative analyses demonstrate that by taking into account contextual relations between pixel segmentation regions within a second-degree neighborhood, spurious label assignments are filtered out, leading to highly accurate semantic segmentations for outdoor scenes.

Vast quantities of videos are now being captured at astonishing rates, but the majority of these are not labelled. To cope with such data, we consider the task of content-based activity recognition in videos without any manually labelled examples, also known as zero-shot video recognition. To achieve this, videos are represented in terms of detected visual concepts, which are then scored as relevant or irrelevant according to their similarity with a given textual query. In this paper, we propose a more robust approach for scoring concepts in order to alleviate many of the brittleness and low precision problems of previous work. Not only do we jointly consider semantic relatedness, visual reliability, and discriminative power. To handle noise and non-linearities in the ranking scores of the selected concepts, we propose a novel pairwise order matrix approach for score aggregation. Extensive experiments on the large-scale TRECVID Multimedia Event Detection data show the superiority of our approach.

In this paper, we focus on automatically detecting events in unconstrained videos without the use of any visual training exemplars. In principle, zero-shot learning makes it possible to train an event detection model based on the assumption that events (e.g. birthday party) can be described by multiple mid-level semantic concepts (e.g. ``blowing candle'', ``birthday cake''). Towards this goal, we first pre-train a bundle of concept classifiers using data from other sources. Then we evaluate the semantic correlation of each concept w.r.t. the event of interest and pick up the relevant concept classifiers, which are applied on all test videos to get multiple prediction score vectors. While most existing systems combine the predictions of the concept classifiers with fixed weights, we propose to learn the optimal weights of the concept classifiers for each testing video by exploring a set of online available videos with free-form text descriptions of their content. To validate the effectiveness of the proposed approach, we have conducted extensive experiments on the latest TRECVID MEDTest 2014, MEDTest 2013 and CCV dataset. The experimental results confirm the superiority of the proposed approach.

Hashing has been widely applied to approximate nearest neighbor search for large-scale multimedia retrieval. Supervised hashing improves the quality of hash coding by exploiting the semantic similarity on data pairs and has received increasing attention recently. For most existing supervised hashing methods for image retrieval, an image is first represented as a vector of hand-crafted or machine-learned features, then quantized by a separate quantization step that generates binary codes. However, suboptimal hash coding may be produced, since the quantization error is not statistically minimized and the feature representation is not optimally compatible with the hash coding. In this paper, we propose a novel Deep Quantization Network (DQN) architecture for supervised hashing, which learns image representation for hash coding and formally control the quantization error. The DQN model constitutes four key components: (1) a sub-network with multiple convolution-pooling layers to capture deep image representations; (2) a fully connected bottleneck layer to generate dimension-reduced representation optimal for hash coding; (3) a pairwise cosine loss layer for similarity-preserving learning; and (4) a product quantization loss for controlling hashing quality and the quantizability of bottleneck representation. Extensive experiments on standard image retrieval datasets show the proposed DQN model yields substantial boosts over latest state-of-the-art hashing methods.

Many planning methods rely on the use of an immediate reward function as a portable and succinct representation of desired behavior. Rewards are often inferred from demonstrated behavior that is assumed to be near-optimal. We examine a framework, Distance Minimization IRL (DM-IRL), for learning reward functions from scores an expert assigns to possibly suboptimal demonstrations. By changing the expert’s role from a demonstrator to a judge, DM-IRL relaxes some of the assumptions present in IRL, enabling learning from the scoring of arbitrary demonstration trajectories with unknown transition functions. DM-IRL complements existing IRL approaches by addressing different assumptions about the expert. We show that DM-IRL is robust to expert scoring error and prove that finding a policy that produces maximally informative trajectories for an expert to score is strongly NP-hard. Experimentally, we demonstrate that the reward function DM-IRL learns from an MDP with an unknown transition model can transfer to an agent with known characteristics in a novel environment, and we achieve successful learning with limited available training data.

The maximum likelihood estimator (MLE) is generally asymptotically consistent but is susceptible to over-fitting. To combat this problem, regularization methods which reduce the variance at the cost of (slightly) increasing the bias are often employed in practice. In this paper, we present an alternative variance reduction (regularization) technique that quantizes the MLE estimates as a post processing step, yielding a smoother model having several tied parameters. We provide and prove error bounds for our new technique and demonstrate experimentally that it often yields models having higher test-set log-likelihood than the ones learned using the MLE. We also propose a new importance sampling algorithm for fast approximate inference in models having several tied parameters. Our experiments show that our new inference algorithm is superior to existing approaches such as Gibbs sampling and MC-SAT on models having tied parameters, learned using our quantization-based approach.