ICCV 2015 论文列表

This paper proposes a novel and more accurate iris segmentation framework to automatically segment iris region from the face images acquired with relaxed imaging under visible or near-infrared illumination, which provides strong feasibility for applications in surveillance, forensics and the search for missing children, etc. The proposed framework is built on a novel total-variation based formulation which uses l1 norm regularization to robustly suppress noisy texture pixels for the accurate iris localization. A series of novel and robust post processing operations are introduced to more accurately localize the limbic boundaries. Our experimental results on three publicly available databases, i.e., FRGC, UBIRIS.v2 and CASIA.v4-distance, achieve significant performance improvement in terms of iris segmentation accuracy over the state-of-the-art approaches in the literature. Besides, we have shown that using iris masks generated from the proposed approach helps to improve iris recognition performance as well. Unlike prior work, all the implementations in this paper are made publicly available to further advance research and applications in biometrics at-d-distance.

We consider the problem of enriching current object detection systems with veridical object sizes and relative depth estimates from a single image. There are several technical challenges to this, such as occlusions, lack of calibration data and the scale ambiguity between object size and distance. These have not been addressed in full generality in previous work. Here we propose to tackle these issues by building upon advances in object recognition and using recently created large-scale datasets. We first introduce the task of amodal bounding box completion, which aims to infer the the full extent of the object instances in the image. We then propose a probabilistic framework for learning category-specific object size distributions from available annotations and leverage these in conjunction with amodal completions to infer veridical sizes of objects in novel images. Finally, we introduce a focal length prediction approach that exploits scene recognition to overcome inherent scale ambiguities and demonstrate qualitative results on challenging real-world scenes.

This paper addresses the problem of unsupervised learning of binary hash codes for efficient cross-modal retrieval. Many unimodal hashing studies have proven that both similarity preservation of data and maintenance of quantization quality are essential for improving retrieval performance with binary hash codes. However, most existing cross-modal hashing methods mainly have focused on the former, and the latter still remains almost untouched. We propose a method to minimize the binary quantization errors, which is tailored to cross-modal hashing. Our approach, named Alternating Co-Quantization (ACQ), alternately seeks binary quantizers for each modality space with the help of connections to other modality data so that they give minimal quantization errors while preserving data similarities. ACQ can be coupled with various existing cross-modal dimension reduction methods such as Canonical Correlation Analysis (CCA) and substantially boosts their retrieval performance in the Hamming space. Extensive experiments demonstrate that ACQ can outperform several state-of-the-art methods, even when it is combined with simple CCA.

Several recent works have shown that image descriptors produced by deep convolutional neural networks provide state-of-the-art performance for image classification and retrieval problems. It also has been shown that the activations from the convolutional layers can be interpreted as local features describing particular image regions. These local features can be aggregated using aggregating methods developed for local features (e.g. Fisher vectors), thus providing new powerful global descriptor. In this paper we investigate possible ways to aggregate local deep features to produce compact descriptors for image retrieval. First, we show that deep features and traditional hand-engineered features have quite different distributions of pairwise similarities, hence existing aggregation methods have to be carefully re-evaluated. Such re-evaluation reveals that in contrast to shallow features, the simple aggregation method based on sum pooling provides the best performance for deep convolutional features. This method is efficient, has few parameters, and bears little risk of overfitting when e.g. learning the PCA matrix. In addition, we suggest a simple yet efficient query expansion scheme suitable for the proposed aggregation method. Overall, the new compact global descriptor improves the state-of-the-art on four common benchmarks considerably.

In this paper, we present a concise framework to approximately construct feature maps for nonlinear additive kernels such as the Intersection, Hellinger's, and Chi^2 kernels. The core idea is to construct for each individual feature a set of anchor points and assign to every query the feature map of its nearest neighbor or the weighted combination of those of its k-nearest neighbors in the anchors. The resultant feature maps can be compactly stored by a group of nearest neighbor (binary) indication vectors along with the anchor feature maps. The approximation error of such an anchored feature mapping approach is analyzed. We evaluate the performance of our approach on large-scale nonlinear support vector machines (SVMs) learning tasks in the context of visual object classification. Experimental results on several benchmark data sets show the superiority of our method over existing feature mapping methods in achieving reasonable trade-off between training time and testing accuracy.

Semi-supervised dictionary learning aims to construct a dictionary by utilizing both labeled and unlabeled data. To enhance the discriminative capability of the learned dictionary, numerous discriminative terms have been proposed by evaluating either the prediction loss or the class separation criterion on the coding vectors of labeled data, but with rare consideration of the power of the coding vectors corresponding to unlabeled data. In this paper, we present a novel semi-supervised dictionary learning method, which uses the informative coding vectors of both labeled and unlabeled data, and adaptively emphasizes the high confidence coding vectors of unlabeled data to enhance the dictionary discriminative capability simultaneously. By doing so, we integrate the discrimination of dictionary, the induction of classifier to new testing data and the transduction of labels to unlabeled data into a unified framework. To solve the proposed problem, an effective iterative algorithm is designed. Experimental results on a series of benchmark databases show that our method outperforms other state-of-the-art dictionary learning methods in most cases.

Hyperspectral imaging is beneficial in a diverse range of applications from diagnostic medicine, to agriculture, to surveillance to name a few. However, hyperspectral images often times suffer from degradation due to the limited light, which introduces noise into the imaging process. In this paper, we propose an effective model for hyperspectral image (HSI) denoising that considers underlying characteristics of HSIs: sparsity across the spatial-spectral domain, high correlation across spectra, and non-local self-similarity over space. We first exploit high correlation across spectra and non-local self-similarity over space in the noisy HSI to learn an adaptive spatial-spectral dictionary. Then, we employ the local and non-local sparsity of the HSI under the learned spatial-spectral dictionary to design an HSI denoising model, which can be effectively solved by an iterative numerical algorithm with parameters that are adaptively adjusted for different clusters and different noise levels. Experimental results on HSI denoising show that the proposed method can provide substantial improvements over the current state-of-the-art HSI denoising methods in terms of both objective metric and subjective visual quality.

With the staggering growth in image and video datasets, algorithms that provide fast similarity search and compact storage are crucial. Hashing methods that map the data into Hamming space have shown promise; however, many of these methods employ a batch-learning strategy in which the computational cost and memory requirements may become intractable and infeasible with larger and larger datasets. To overcome these challenges, we propose an online learning algorithm based on stochastic gradient descent in which the hash functions are updated iteratively with streaming data. In experiments with three image retrieval benchmarks, our online algorithm attains retrieval accuracy that is comparable to competing state-of-the-art batch-learning solutions, while our formulation is orders of magnitude faster and being online it is adaptable to the variations of the data. Moreover, our formulation yields improved retrieval performance over a recently reported online hashing technique, Online Kernel Hashing.

We propose an efficient online video filtering method, called adaptive exponential filtering (AES) to refine pixel prediction maps. Assuming each pixel is associated with a discriminative prediction score, the proposed AES applies exponentially decreasing weights over time to smooth the prediction score of each pixel, similar to classic exponential smoothing. However, instead of fixing the spatial pixel location to perform temporal filtering, we trace each pixel in the past frames by finding the optimal path that can bring the maximum exponential smoothing score, thus performing adaptive and non-linear filtering. Thanks to the pixel tracing, AES can better address object movements and avoid over-smoothing. To enable real-time filtering, we propose a linear-complexity dynamic programming scheme that can trace all pixels simultaneously. We apply the proposed filtering method to improve both saliency detection maps and scene parsing maps. The comparisons with average and exponential filtering, as well as state-of-the-art methods, validate that our AES can effectively refine the pixel prediction maps, without using the original video again.

Hough voting in a geometric transformation space allows us to realize spatial verification, but remains sensitive to feature detection errors because of the inflexible quantization of single feature correspondences. To handle this problem, we propose a new method, called adaptive dither voting, for robust spatial verification. For each correspondence, instead of hard-mapping it to a single transformation, the method augments its description by using multiple dithered transformations that are deterministically generated by the other correspondences. The method reduces the probability of losing correspondences during transformation quantization, and provides high robustness as regards mismatches by imposing three geometric constraints on the dithering process. We also propose exploiting the non-uniformity of a Hough histogram as the spatial similarity to handle multiple matching surfaces. Extensive experiments conducted on four datasets show the superiority of our method. The method outperforms its state-of-the-art counterparts in both accuracy and scalability, especially when it comes to the retrieval of small, rotated objects.

In this paper, we propose an activity auto-completion (AAC) model for human activity prediction by formulating activity prediction as a query auto-completion (QAC) problem in information retrieval. First, we extract discriminative patches in frames of videos. A video is represented based on these patches and divided into a collection of segments, each of which is regarded as a character typed in the search box. Then a partially observed video is considered as an activity prefix, consisting of one or more characters. Finally, the missing observation of an activity is predicted as the activity candidates provided by the auto-completion model. The candidates are matched against the activity prefix on-the-fly and ranked by a learning-to-rank algorithm. We validate our method on UT-Interaction Set #1 and Set #2 [19]. The experimental results show that the proposed activity auto-completion model achieves promising performance.

How can we reuse existing knowledge, in the form of available datasets, when solving a new and apparently unrelated target task from a set of unlabeled data? In this work we make a first contribution to answer this question in the context of image classification. We frame this quest as an active learning problem and use zero-shot classifiers to guide the learning process by linking the new task to the the existing classifiers. By revisiting the dual formulation of adaptive SVM, we reveal two basic conditions to choose greedily only the most relevant samples to be annotated. On this basis we propose an effective active learning algorithm which learns the best possible target classification model with minimum human labeling effort. Extensive experiments on two challenging datasets show the value of our approach compared to the state-of-the-art active learning methodologies, as well as its potential to reuse past datasets with minimal effort for future tasks.

The central projection model commonly used to model cameras as well as projectors, results in similar advantages and disadvantages in both types of system. Considering the case of active stereo systems using a projector and camera setup, a central projection model creates several problems; among them, narrow depth range and necessity of wide baseline are crucial. In the paper, we solve the problems by introducing a light field projector, which can project a depth-dependent pattern. The light field projector is realized by attaching a coded aperture with a high frequency mask in front of the lens of the video projector, which also projects a high frequency pattern. Because the light field projector cannot be approximated by a thin lens model and a precise calibration method is not established yet, an image-based approach is proposed to apply a stereo technique to the system. Although image-based techniques usually require a large database and often imply heavy computational costs, we propose a hierarchical approach and a feature-based search for solution. In the experiments, it is confirmed that our method can accurately recover the dense shape of curved and textured objects for a wide range of depths from a single captured image.

We present an active detection model for localizing objects in scenes. The model is class-specific and allows an agent to focus attention on candidate regions for identifying the correct location of a target object. This agent learns to deform a bounding box using simple transformation actions, with the goal of determining the most specific location of target objects following top-down reasoning. The proposed localization agent is trained using deep reinforcement learning, and evaluated on the Pascal VOC 2007 dataset. We show that agents guided by the proposed model are able to localize a single instance of an object after analyzing only between 11 and 25 regions in an image, and obtain the best detection results among systems that do not use object proposals for object localization.

We investigate the importance of parts for the tasks of action and attribute classification. We develop a part-based approach by leveraging convolutional network features inspired by recent advances in computer vision. Our part detectors are a deep version of poselets and capture parts of the human body under a distinct set of poses. For the tasks of action and attribute classification, we train holistic convolutional neural networks and show that adding parts leads to top-performing results for both tasks. We observe that for deeper networks parts are less significant. In addition, we demonstrate the effectiveness of our approach when we replace an oracle person detector, as is the default in the current evaluation protocol for both tasks, with a state-of-the-art person detection system.

We elicit from a fundamental definition of action low-level attributes that can reveal agency and intentionality. These descriptors are mainly trajectory-based, measuring sudden changes, temporal synchrony, and repetitiveness. The actionness map can be used to localize actions in a way that is generic across action and agent types. Furthermore, it also groups interacting regions into a useful unit of analysis, which is crucial for recognition of actions involving interactions. We then implement an actionness-driven pooling scheme to improve action recognition performance. Experimental results on three datasets show the advantages of our method on both action detection and action recognition comparing with other state-of-the-art methods.

Realistic videos of human actions exhibit rich spatiotemporal structures at multiple levels of granularity: an action can always be decomposed into multiple finer-grained elements in both space and time. To capture this intuition, we propose to represent videos by a hierarchy of mid-level action elements (MAEs), where each MAE corresponds to an action-related spatiotemporal segment in the video. We introduce an unsupervised method to generate this representation from videos. Our method is capable of distinguishing action-related segments from background segments and representing actions at multiple spatiotemporal resolutions. Given a set of spatiotemporal segments generated from the training data, we introduce a discriminative clustering algorithm that automatically discovers MAEs at multiple levels of granularity. We develop structured models that capture a rich set of spatial, temporal and hierarchical relations among the segments, where the action label and multiple levels of MAE labels are jointly inferred. The proposed model achieves state-of-the-art performance in multiple action recognition benchmarks. Moreover, we demonstrate the effectiveness of our model in real-world applications such as action recognition in large-scale untrimmed videos and action parsing.

This paper presents an efficient approach for localizing actions by learning contextual relations, in the form of relative locations between different video regions. We begin by over-segmenting the videos into supervoxels, which have the ability to preserve action boundaries and also reduce the complexity of the problem. Context relations are learned during training which capture displacements from all the supervoxels in a video to those belonging to foreground actions. Then, given a testing video, we select a supervoxel randomly and use the context information acquired during training to estimate the probability of each supervoxel belonging to the foreground action. The walk proceeds to a new supervoxel and the process is repeated for a few steps. This ``context walk'' generates a conditional distribution of an action over all the supervoxels. A Conditional Random Field is then used to find action proposals in the video, whose confidences are obtained using SVMs. We validated the proposed approach on several datasets and show that context in the form of relative displacements between supervoxels can be extremely useful for action localization. This also results in significantly fewer evaluations of the classifier, in sharp contrast to the alternate sliding window approaches.

Explicitly using human detection and pose estimation has found limited success in action recognition problems. This may be due to the complexity in the articulated motion human exhibit. Yet, we know that action requires an actor and intention. This paper hence seeks to understand the spatiotemporal properties of intentional movement and how to capture such intentional movement without relying on challenging human detection and tracking. We conduct a quantitative analysis of intentional movement, and our findings motivate a new approach for implicit intentional movement extraction that is based on spatiotemporal trajectory clustering by leveraging the properties of intentional movement. The intentional movement clusters are then used as action proposals for detection. Our results on three action detection benchmarks indicate the relevance of focusing on intentional movement for action detection; our method significantly outperforms the state of the art on the challenging MSR-II multi-action video benchmark.