CVPR 2017 Papers

Hyperspectral imaging is a useful technique for various computer vision tasks such as material recognition. However, such technique usually requires an expensive and professional setup and is time-consuming because a conventional hyperspectral image consists of a large number of observations. In this paper, we propose a novel technique of one-shot hyperspectral imaging using faced reflectors on which color filters are attached. The key idea is based on the principle that each of multiple reflections on the filters has a different spectrum, which allows us to observe multiple intensities through different spectra. Our technique can be implemented either by a coupled mirror or a kaleidoscope geometry. Experimental results show that our technique is capable of accurately capturing a hyperspectral image by using a coupled mirror setup which is readily available.

Camera relocalisation is an important problem in computer vision, with applications in simultaneous localisation and mapping, virtual/augmented reality and navigation. Common techniques either match the current image against keyframes with known poses coming from a tracker, or establish 2D-to-3D correspondences between keypoints in the current image and points in the scene in order to estimate the camera pose. Recently, regression forests have become a popular alternative to establish such correspondences. They achieve accurate results, but must be trained offline on the target scene, preventing relocalisation in new environments. In this paper, we show how to circumvent this limitation by adapting a pre-trained forest to a new scene on the fly. Our adapted forests achieve relocalisation performance that is on par with that of offline forests, and our approach runs in under 150ms, making it desirable for real-time systems that require online relocalisation.

We present new methods of simultaneously estimating camera geometry and time shift from video sequences from multiple unsynchronized cameras. Algorithms for simultaneous computation of a fundamental matrix or a homography with unknown time shift between images are developed. Our methods use minimal correspondence sets (eight for fundamental matrix and four and a half for homography) and therefore are suitable for robust estimation using RANSAC. Furthermore, we present an iterative algorithm that extends the applicability on sequences which are significantly unsynchronized, finding the correct time shift up to several seconds. We evaluated the methods on synthetic and wide range of real world datasets and the results show a broad applicability to the problem of camera synchronization.

Blind deconvolution is the problem of recovering a convolutional kernel and an activation signal from their convolution. This problem is ill-posed without further constraints or priors. This paper studies the situation where the nonzero entries in the activation signal are sparsely and randomly populated.. We normalize the convolution kernel to have unit Frobenius norm and cast the sparse blind deconvolution problem as a nonconvex optimization problem over the sphere. With this spherical constraint, every spurious local minimum turns out to be close to some signed shift truncation of the ground truth, under certain hypotheses. This benign property motivates an effective two stage algorithm that recovers the ground truth from the partial information offered by a suboptimal local minimum. This geometry-inspired algorithm recovers the ground truth for certain microscopy problems, also exhibits promising performance in the more challenging image deblurring problem. Our insights into the global geometry and the two stage algorithm extend to the convolutional dictionary learning problem, where a superposition of multiple convolution signals is observed.

Visible watermarking is a widely-used technique for marking and protecting copyrights of many millions of images on the web, yet it suffers from an inherent security flaw---watermarks are typically added in a consistent manner to many images. We show that this consistency allows to automatically estimate the watermark and recover the original images with high accuracy. Specifically, we present a generalized multi-image matting algorithm that takes a watermarked image collection as input and automatically estimates the "foreground" (watermark), its alpha matte, and the "background" (original) images. Since such an attack relies on the consistency of watermarks across image collection, we explore and evaluate how it is affected by various types of inconsistencies in the watermark embedding that could potentially be used to make watermarking more secured. We demonstrate the algorithm on stock imagery available on the web, and provide extensive quantitative analysis on synthetic watermarked data. A key takeaway message of this paper is that visible watermarks should be designed to not only be robust against removal from a single image, but to be more resistant to mass-scale removal from image collections as well.

Human motion modelling is a classical problem at the intersection of graphics and computer vision, with applications spanning human-computer interaction, motion synthesis, and motion prediction for virtual and augmented reality. Following the success of deep learning methods in several computer vision tasks, recent work has focused on using deep recurrent neural networks (RNNs) to model human motion, with the goal of learning time-dependent representations that perform tasks such as short-term motion prediction and long-term human motion synthesis. We examine recent work, with a focus on the evaluation methodologies commonly used in the literature, and show that, surprisingly, state of the art performance can be achieved by a simple baseline that does not attempt to model motion at all. We investigate this result, and analyze recent RNN methods by looking at the architectures, loss functions, and training procedures used in state-of-the-art approaches. We propose three changes to the standard RNN models typically used for human motion, which results in a simple and scalable RNN architecture that obtains state-of-the-art performance on human motion prediction.

Deep compression refers to removing the redundancy of parameters and feature maps for deep learning models. Low-rank approximation and pruning for sparse structures play a vital role in many compression works. However, weight filters tend to be both low-rank and sparse. Neglecting either part of these structure information in previous methods results in iteratively retraining, compromising accuracy, and low compression rates. Here we propose a unified framework integrating the low-rank and sparse decomposition of weight matrices with the feature map reconstructions. Our model includes methods like pruning connections as special cases, and is optimized by a fast SVD-free algorithm. It has been theoretically proven that, with a small sample, due to its generalizability, our model can well reconstruct the feature maps on both training and test data, which results in less compromising accuracy prior to the subsequent retraining. With such a "warm start" to retrain, the compression method always possesses several merits: (a) higher compression rates, (b) little loss of accuracy, and (c) fewer rounds to compress deep models. The experimental results on several popular models such as AlexNet, VGG-16, and GoogLeNet show that our model can significantly reduce the parameters for both convolutional and fully-connected layers. As a result, our model reduces the size of VGG-16 by 15x, better than other recent compression methods that use a single strategy.

We present OctNet, a representation for deep learning with sparse 3D data. In contrast to existing models, our representation enables 3D convolutional networks which are both deep and high resolution. Towards this goal, we exploit the sparsity in the input data to hierarchically partition the space using a set of unbalanced octrees where each leaf node stores a pooled feature representation. This allows to focus memory allocation and computation to the relevant dense regions and enables deeper networks without compromising resolution. We demonstrate the utility of our OctNet representation by analyzing the impact of resolution on several 3D tasks including 3D object classification, orientation estimation and point cloud labeling.

This work aims to build pixel-to-pixel correspondences between images from the same visual class but with different geometries and visual similarities. This task is particularly challenging because (i) their visual content is similar only on the high-level structure, and (ii) background clutters keep bringing in noises. To address these problems, this paper proposes an object-aware method to estimate per-pixel correspondences from semantic to low-level by learning a classifier for each selected discriminative grid cell and guiding the localization of every pixel under the semantic constraint. Specifically, an Object-aware Hierarchical Graph (OHG) model is constructed to regulate matching consistency from one coarse grid cell containing whole object(s), to fine grid cells covering smaller semantic elements, and finally to every pixel. A guidance layer is introduced as the semantic constraint on local structure matching. In addition, we propose to learn the important high-level structure for each grid cell in an "objectness-driven" way as an alternative to handcrafted descriptors in defining a better visual similarity. The proposed method has been extensively evaluated on various challenging benchmarks and real-world images. The results show that our method significantly outperforms the state-of-the-arts in terms of semantic flow accuracy.

We investigate a principle way to progressively mine discriminative object regions using classification networks to address the weakly-supervised semantic segmentation problems. Classification networks are only responsive to small and sparse discriminative regions from the object of interest, which deviates from the requirement of the segmentation task that needs to localize dense, interior and integral regions for pixel-wise inference. To mitigate this gap, we propose a new adversarial erasing approach for localizing and expanding object regions progressively. Starting with a single small object region, our proposed approach drives the classification network to sequentially discover new and complement object regions by erasing the current mined regions in an adversarial manner. These localized regions eventually constitute a dense and complete object region for learning semantic segmentation. To further enhance the quality of the discovered regions by adversarial erasing, an online prohibitive segmentation learning approach is developed to collaborate with adversarial erasing by providing auxiliary segmentation supervision modulated by the more reliable classification scores. Despite its apparent simplicity, the proposed approach achieves 55.0% and 55.7% mean Intersection-over-Union (mIoU) scores on PASCAL VOC 2012 val and test sets, which are the new state-of-the-arts.

Object detection in videos has drawn increasing attention recently with the introduction of the large-scale ImageNet VID dataset. Different from object detection in static images, temporal information in videos is vital for object detection. To fully utilize temporal information, state-of-the-art methods are based on spatiotemporal tubelets, which are essentially sequences of associated bounding boxes across time. However, the existing methods have major limitations in generating tubelets in terms of quality and efficiency. Motion-based methods are able to obtain dense tubelets efficiently, but the lengths are generally only several frames, which is not optimal for incorporating long-term temporal information. Appearance-based methods, usually involving generic object tracking, could generate long tubelets, but are usually computationally expensive. In this work, we propose a framework for object detection in videos, which consists of a novel tubelet proposal network to efficiently generate spatiotemporal proposals, and a Long Short-term Memory (LSTM) network that incorporates temporal information from tubelet proposals for achieving high object detection accuracy in videos. Experiments on the large-scale ImageNet VID dataset demonstrate the effectiveness of the proposed framework for object detection in videos.

Recent advances in the joint processing of images have certainly shown its advantages over the individual processing. Different from the existing works geared towards co-segmentation or co-localization, in this paper, we explore a new joint processing topic: co-skeletonization, which is defined as joint skeleton extraction of common objects in a set of semantically similar images. Object skeletonization in real world images is a challenging problem, because there is no prior knowledge of the object's shape if we consider only a single image. This motivates us to resort to the idea of object co-skeletonization hoping that the commonness prior existing across the similar images may help, just as it does for other joint processing problems such as co-segmentation. Noting that skeleton can provide good scribbles for segmentation, and skeletonization, in turn, needs good segmentation, we propose a coupled framework for co-skeletonization and co-segmentation tasks so that they are well informed by each other, and benefit each other synergistically. Since it is a new problem, we also construct a benchmark dataset for the co-skeletonization task. Extensive experiments demonstrate that proposed method achieves very competitive results.

We propose a novel deep layer cascade (LC) method to improve the accuracy and speed of semantic segmentation. Unlike the conventional model cascade (MC) that is composed of multiple independent models, LC treats a single deep model as a cascade of several sub-models. Earlier sub-models are trained to handle easy and confident regions, and they progressively feed-forward harder regions to the next sub-model for processing. Convolutions are only calculated on these regions to reduce computations. The proposed method possesses several advantages. First, LC classifies most of the easy regions in the shallow stage and makes deeper stage focuses on a few hard regions. Such an adaptive and 'difficulty-aware' learning improves segmentation performance. Second, LC accelerates both training and testing of deep network thanks to early decisions in the shallow stage. Third, in comparison to MC, LC is an end-to-end trainable framework, allowing joint learning of all sub-models. We evaluate our method on PASCAL VOC and Cityscapes datasets, achieving state-of-the-art performance and fast speed.

Surveillance cameras today often capture NIR (near infrared) images in low-light environments. However, most face datasets accessible for training and verification are only collected in the VIS (visible light) spectrum. It remains a challenging problem to match NIR to VIS face images due to the different light spectrum. Recently, breakthroughs have been made for VIS face recognition by applying deep learning on a huge amount of labeled VIS face samples. The same deep learning approach cannot be simply applied to NIR face recognition for two main reasons: First, much limited NIR face images are available for training compared to the VIS spectrum. Second, face galleries to be matched are mostly available only in the VIS spectrum. In this paper, we propose an approach to extend the deep learning breakthrough for VIS face recognition to the NIR spectrum, without retraining the underlying deep models that see only VIS faces. Our approach consists of two core components, cross-spectral hallucination and low-rank embedding, to optimize respectively input and output of a VIS deep model for cross-spectral face recognition. Cross-spectral hallucination produces VIS faces from NIR images through a deep learning approach. Low-rank embedding restores a low-rank structure for faces deep features across both NIR and VIS spectrum. We observe that it is often equally effective to perform hallucination to input NIR images or low-rank embedding to output deep features for a VIS deep model for cross-spectral recognition. When hallucination and low-rank embedding are deployed together, we observe significant further improvement; we obtain state-of-the-art accuracy on the CASIA NIR-VIS v2.0 benchmark, without the need at all to re-train the recognition system.

In this work, we introduce a highly efficient algorithm to address the nonnegative matrix underapproximation (NMU) problem, i.e., nonnegative matrix factorization (NMF) with an additional underapproximation constraint. NMU results are interesting as, compared to traditional NMF, they present additional sparsity and part-based behavior, explaining unique data features. To show these features in practice, we first present an application to the analysis of climate data. We then present an NMU-based algorithm to robustly fit multiple parametric models to a dataset. The proposed approach delivers state-of-the-art results for the estimation of multiple fundamental matrices and homographies, outperforming other alternatives in the literature and exemplifying the use of efficient NMU computations.

Several works have shown that relationships between data points (i.e., context) in structured data can be exploited to obtain better recognition performance. In this paper, we explore a different, but related, problem: how can these inter-relationships be used to efficiently learn and continuously update a recognition model, with minimal human labeling effort. Towards this goal, we propose an active learning framework to select an optimal subset of data points for manual labeling by exploiting the relationships between them. We construct a graph from the unlabeled data to represent the underlying structure, such that each node represents a data point, and edges represent the inter-relationships between them. Thereafter, considering the flow of beliefs in this graph, we choose those samples for labeling which minimize the joint entropy of the nodes of the graph. This results in significant reduction in manual labeling effort without compromising recognition performance. Our method chooses non-uniform number of samples from each batch of streaming data depending on its information content. Also, the submodular property of our objective function makes it computationally efficient to optimize. The proposed framework is demonstrated in various applications, including document analysis, scene-object recognition, and activity recognition.

Saliency detection aims to highlight the most relevant objects in an image. Methods using conventional models struggle whenever salient objects are pictured on top of a cluttered background while deep neural nets suffer from excess complexity and slow evaluation speeds. In this paper, we propose a simplified convolutional neural network which combines local and global information through a multi-resolution 4x5 grid structure. Instead of enforcing spacial coherence with a CRF or superpixels as is usually the case, we implemented a loss function inspired by the Mumford-Shah functional which penalizes errors on the boundary. We trained our model on the MSRA-B dataset, and tested it on six different saliency benchmark datasets. Results show that our method is on par with the state-of-the-art while reducing computation time by a factor of 18 to 100 times, enabling near real-time, high performance saliency detection.

We propose a novel deep network architecture for grayscale and color image denoising that is based on a non-local image model. Our motivation for the overall design of the proposed network stems from variational methods that exploit the inherent non-local self-similarity property of natural images. We build on this concept and introduce deep networks that perform non-local processing and at the same time they significantly benefit from discriminative learning. Experiments on the Berkeley segmentation dataset, comparing several state-of-the-art methods, show that the proposed non-local models achieve the best reported denoising performance both for grayscale and color images for all the tested noise levels. It is also worth noting that this increase in performance comes at no extra cost on the capacity of the network compared to existing alternative deep network architectures. In addition, we highlight a direct link of the proposed non-local models to convolutional neural networks. This connection is of significant importance since it allows our models to take full advantage of the latest advances on GPU computing in deep learning and makes them amenable to efficient implementations through their inherent parallelism.

Vibration measurement systems are widely used in the industry. A variety of vibration measurement techniques are proposed, including methods using an acceleration sensor, a laser displacement meter, and tracking a marker using a camera. However, these methods have limitations that allow only one point to be measured and require markers. We present a novel, non-contact full field joint measurement technique both of vibrations and shape based on phase-shifting. Our key idea is to acquire the frequency of vibrating objects using FFT to analyze the phase-shift error of vibrating objects. Our proposed algorithm estimates the phase-shift error by iterating frame-to-frame optimization and pixel-to-pixel optimization. A feature of our approach is to measure the surface of vibrating at different frequencies without markers or texture in full fields. Our developed system is a low cost system, which is composed of a digital-light-processing (DLP) projector and camera (100 frames per second). The results of our experiments show that low frequency vibration of objects can be measured with high accuracy in non-contact. Also, reconstruction of the vibrating object surface can be performed with high accuracy.