CVPR 2016 论文列表

We present recursive recurrent neural networks with attention modeling (R2AM) for lexicon-free optical character recognition in natural scene images. The primary advantages of the proposed method are: (1) use of recursive convolutional neural networks (CNNs), which allow for parametrically efficient and effective image feature extraction; (2) an implicitly learned character-level language model, embodied in a recurrent neural network which avoids the need to use N-grams; and (3) the use of a soft-attention mechanism, allowing the model to selectively exploit image features in a coordinated way, and allowing for end-to-end training within a standard backpropagation framework. We validate our method with state-of-the-art performance on challenging benchmark datasets: Street View Text, IIIT5k, ICDAR and Synth90k.

Robust visual tracking is a challenging task in computer vision. Due to the accumulation and propagation of estimation error, model drifting often occurs and degrades the tracking performance. To mitigate this problem, in this paper we propose a novel tracking method called Recurrently Target-attending Tracking (RTT). RTT attempts to identify and exploit those reliable parts which are beneficial for the overall tracking process. To bypass occlusion and discover reliable components, multi-directional Recurrent Neural Networks (RNNs) are employed in RTT to capture long-range contextual cues by traversing a candidate spatial region from multiple directions. The produced confidence maps from the RNNs are employed to adaptively regularize the learning of discriminative correlation filters by suppressing clutter background noises while making full use of the information from reliable parts. To solve the weighted correlation filters, we especially derive an efficient closed-form solution with a sharp reduction in computation complexity. Extensive experiments demonstrate that our proposed RTT is more competitive over those correlation filter based methods.

Modeling the aging process of human face is important for cross-age face verification and recognition. In this paper, we introduce a recurrent face aging (RFA) framework based on a recurrent neural network which can identify the ages of people from 0 to 80. Due to the lack of labeled face data of the same person captured in a long range of ages, traditional face aging models usually split the ages into discrete groups and learn a one-step face feature transformation for each pair of adjacent age groups. However, those methods neglect the in-between evolving states between the adjacent age groups and the synthesized faces often suffer from severe ghosting artifacts. Since human face aging is a smooth progression, it is more appropriate to age the face by going through smooth transition states. In this way, the ghosting artifacts can be effectively eliminated and the intermediate aged faces between two discrete age groups can also be obtained. Towards this target, we employ a two-layer gated recurrent unit as the basic recurrent module whose bottom layer encodes a young face to a latent representation and the top layer decodes the representation to a corresponding older face. The experimental results demonstrate our proposed RFA provides better aging faces over other state-of-the-art age progression methods.

In this paper we propose a novel recurrent neural network architecture for video-based person re-identification. Given the video sequence of a person, features are extracted from each frame using a convolutional neural network that incorporates a recurrent final layer, which allows information to flow between time-steps. The features from all time-steps are then combined using temporal pooling to give an overall appearance feature for the complete sequence. The convolutional network, recurrent layer, and temporal pooling layer, are jointly trained to act as a feature extractor for video-based re-identification using a Siamese network architecture. Our approach makes use of colour and optical flow information in order to capture appearance and motion information which is useful for video re-identification. Experiments are conduced on the iLIDS-VID and PRID-2011 datasets to show that this approach outperforms existing methods of video-based re-identification.

Convolutional-deconvolution networks can be adopted to perform end-to-end saliency detection. But, they do not work well with objects of multiple scales. To overcome such a limitation, in this work, we propose a recurrent attentional convolutional-deconvolution network (RACDNN). Using spatial transformer and recurrent network units, RACDNN is able to iteratively attend to selected image sub-regions to perform saliency refinement progressively. Besides tackling the scale problem, RACDNN can also learn context-aware features from past iterations to enhance saliency refinement in future iterations. Experiments on several challenging saliency detection datasets validate the effectiveness of RACDNN, and show that RACDNN outperforms state-of-the-art saliency detection methods.

We present an attention-based model that reasons on human body shape and motion dynamics to identify individuals in the absence of RGB information, hence in the dark. Our approach leverages unique 4D spatio-temporal signatures to address the identification problem across days. Formulated as a reinforcement learning task, our model is based on a combination of convolutional and recurrent neural networks with the goal of identifying small, discriminative regions indicative of human identity. We demonstrate that our model produces state-of-the-art results on several published datasets given only depth images. We further study the robustness of our model towards viewpoint, appearance, and volumetric changes. Finally, we share insights gleaned from interpretable 2D, 3D, and 4D visualizations of our model's spatio-temporal attention.

Collecting training images for all visual categories is not only expensive but also impractical. Zero-shot learning (ZSL), especially using attributes, offers a pragmatic solution to this problem. However, at test time most attribute-based methods require a full description of attribute associations for each unseen class. Providing these associations is time consuming and often requires domain specific knowledge. In this work, we aim to carry out attribute-based zero-shot classification in an unsupervised manner. We propose an approach to learn relations that couples class embeddings with their corresponding attributes. Given only the name of an unseen class, the learned relationship model is used to automatically predict the class-attribute associations. Furthermore, our model facilitates transferring attributes across data sets without additional effort. Integrating knowledge from multiple sources results in a significant additional improvement in performance. We evaluate on two public data sets: Animals with Attributes and aPascal/aYahoo. Our approach outperforms state-of-the-art methods in both predicting class-attribute associations and unsupervised ZSL by a large margin.

This paper presents a method for recovering shape and normal of a transparent object from a single viewpoint using a Time-of-Flight (ToF) camera. Our method is built upon the fact that the speed of light varies with the refractive index of the medium and therefore the depth measurement of a transparent object with a ToF camera may be distorted. We show that, from this ToF distortion, the refractive light path can be uniquely determined by estimating a single parameter. We estimate this parameter by introducing a surface normal consistency between the one determined by a light path candidate and the other computed from the corresponding shape. The proposed method is evaluated by both simulation and real-world experiments and shows faithful transparent shape recovery.

Object detection and 6D pose estimation in the crowd (scenes with multiple object instances, severe foreground occlusions and background distractors), has become an important problem in many rapidly evolving technological areas such as robotics and augmented reality. Single shot-based 6D pose estimators with manually designed features are still unable to tackle the above challenges, motivating the research towards unsupervised feature learning and next-best-view estimation. In this work, we present a complete framework for both single shot-based 6D object pose estimation and next-best-view prediction based on Hough Forests, the state of the art object pose estimator that performs classification and regression jointly. Rather than using manually designed features we a) propose an unsupervised feature learnt from depth-invariant patches using a Sparse Autoencoder and b) offer an extensive evaluation of various state of the art features. Furthermore, taking advantage of the clustering performed in the leaf nodes of Hough Forests, we learn to estimate the reduction of uncertainty in other views, formulating the problem of selecting the next-best-view. To further improve pose estimation, we propose an improved joint registration and hypotheses verification module as a final refinement step to reject false detections. We provide two additional challenging datasets inspired from realistic scenarios to extensively evaluate the state of the art and our framework. One is related to domestic environments and the other depicts a bin-picking scenario mostly found in industrial settings. We show that our framework significantly outperforms state of the art both on public and on our datasets.

Reconstruction of shapes and appearances of thin film objects can be applied to many fields such as industrial inspection, biological analysis, and archeology research. However, it comes with many challenging issues because the appearances of thin film can change dramatically depending on view and light directions. The appearance is deeply dependent on not only the shapes but also the optical parameters of thin film. In this paper, we propose a novel method to estimate shapes and film thickness. First, we narrow down candidates of zenith angle by degree of polarization and determine it by the intensity of thin film which increases monotonically along the zenith angle. Second, we determine azimuth angle from occluding boundaries. Finally, we estimate the film thickness by comparing a look-up table of color along the thickness and zenith angle with captured images. We experimentally evaluated the accuracy of estimated shapes and appearances and found that our proposed method is effective.

The goal of this paper is to present a non-iterative and more importantly an extremely fast algorithm to reconstruct images from compressively sensed (CS) random measurements. To this end, we propose a novel convolutional neural network (CNN) architecture which takes in CS measurements of an image as input and outputs an intermediate reconstruction. We call this network, ReconNet. The intermediate reconstruction is fed into an off-the-shelf denoiser to obtain the final reconstructed image. On a standard dataset of images we show significant improvements in reconstruction results (both in terms of PSNR and time complexity) over state-of-the-art iterative CS reconstruction algorithms at various measurement rates. Further, through qualitative experiments on real data collected using our block SPC (single pixel camera), we show that our network is highly robust to sensor noise and can recover visually better quality images than competitive algorithms at extremely low sensing rates of 0.1 and 0.04. To demonstrate that our algorithm can recover semantically informative images even at a low measurement rate of 0.01, we present a very robust proof of concept real-time visual tracking application.

Deep neural networks with alternating convolutional, max-pooling and decimation layers are widely used in state of the art architectures for computer vision. Max-pooling purposefully discards precise spatial information in order to create features that are more robust, and typically organized as lower resolution spatial feature maps. On some tasks, such as whole-image classification, max-pooling derived features are well suited; however, for tasks requiring precise localization, such as pixel level prediction and segmentation, max-pooling destroys exactly the information required to perform well. Precise localization may be preserved by shallow convnets without pooling but at the expense of robustness. Can we have our max-pooled multi-layered cake and eat it too? Several papers have proposed summation and concatenation based methods for combining upsampled coarse, abstract features with finer features to produce robust pixel level predictions. Here we introduce another model --- dubbed Recombinator Networks --- where coarse features inform finer features early in their formation such that finer features can make use of several layers of computation in deciding how to use coarse features. The model is trained once, end-to-end and performs better than summation-based architectures, reducing the error from the previous state of the art on two facial keypoint datasets, AFW and AFLW, by 30% and beating the current state-of-the-art on 300W without using extra data. We improve performance even further by adding a denoising prediction model based on a novel convnet formulation.

We aim to understand the dynamics of social interactions between two people by recognizing their actions and reactions using a head-mounted camera. Our work will impact several first-person vision tasks that need the detailed understanding of social interactions, such as automatic video summarization of group events and assistive systems. To recognize micro-level actions and reactions, such as slight shifts in attention, subtle nodding, or small hand actions, where only subtle body motion is apparent, we propose to use paired egocentric videos recorded by two interacting people. We show that the first-person and second-person points-of-view features of two people, enabled by paired egocentric videos, are complementary and essential for reliably recognizing micro-actions and reactions. We also build a new dataset of dyadic (two-persons) interactions that comprises more than 1000 pairs of egocentric videos to enable systematic evaluations on the task of micro-action and reaction recognition.

Advanced computer vision and machine learning techniques tried to automatically categorize the emotions elicited by abstract paintings with limited success. Since the annotation of the emotional content is highly resource-consuming, datasets of abstract paintings are either constrained in size or partially annotated. Consequently, it is natural to address the targeted task within a transductive framework. Intuitively, the use of multi-label classification techniques is desirable so to synergically exploit the relations between multiple latent variables, such as emotional content, technique, author, etc. A very popular approach for transductive multi-label recognition under linear classification settings is matrix completion. In this study we introduce non-linear matrix completion (NLMC), thus extending classical linear matrix completion techniques to the non-linear case. Together with the theory grounding the model, we propose an efficient optimization solver. As shown by our extensive experimental validation on two publicly available datasets, NLMC outperforms state-of-the-art methods when recognizing emotions from abstract paintings.

Physical fluents, a term originally used by Newton [40], refers to time-varying object states in dynamic scenes. In this paper, we are interested in inferring the fluents of vehicles from video. For example, a door (hood, trunk) is open or closed through various actions, light is blinking to turn. Recognizing these fluents has broad applications, yet have received scant attention in the computer vision literature. Car fluent recognition entails a unified framework for car detection, car part localization and part status recognition, which is made difficult by large structural and appearance variations, low resolutions and occlusions. This paper learns a spatial-temporal And-Or hierarchical model to represent car fluents. The learning of this model is formulated under the latent structural SVM framework. Since there are no publicly related dataset, we collect and annotate a car fluent dataset consisting of car videos with diverse fluents. In experiments, the proposed method outperforms several highly related baseline methods in terms of car fluent recognition and car part localization.

We present a novel dataset and a novel algorithm for recognizing activities of daily living (ADL) from a first-person wearable camera. Handled objects are crucially important for egocentric ADL recognition. For specific examination of objects related to users' actions separately from other objects in an environment, many previous works have addressed the detection of handled objects in images captured from head-mounted and chest-mounted cameras. Nevertheless, detecting handled objects is not always easy because they tend to appear small in images. They can be occluded by a user's body. As described herein, we mount a camera on a user's wrist. A wrist-mounted camera can capture handled objects at a large scale, and thus it enables us to skip the object detection process. To compare a wrist-mounted camera and a head-mounted camera, we also developed a novel and publicly available dataset that includes videos and annotations of daily activities captured simultaneously by both cameras. Additionally, we propose a discriminative video representation that retains spatial and temporal information after encoding the frame descriptors extracted by convolutional neural networks.

Recently, several models based on deep neural networks have achieved great success in terms of both reconstruction accuracy and computational performance for single image super-resolution. In these methods, the low resolution (LR) input image is upscaled to the high resolution (HR) space using a single filter, commonly bicubic interpolation, before reconstruction. This means that the super-resolution (SR) operation is performed in HR space. We demonstrate that this is sub-optimal and adds computational complexity. In this paper, we present the first convolutional neural network (CNN) capable of real-time SR of 1080p videos on a single K2 GPU. To achieve this, we propose a novel CNN architecture where the feature maps are extracted in the LR space. In addition, we introduce an efficient sub-pixel convolution layer which learns an array of upscaling filters to upscale the final LR feature maps into the HR output. By doing so, we effectively replace the handcrafted bicubic filter in the SR pipeline with more complex upscaling filters specifically trained for each feature map, whilst also reducing the computational complexity of the overall SR operation. We evaluate the proposed approach using images and videos from publicly available datasets and show that it performs significantly better (+0.15dB on Images and +0.39dB on Videos) and is an order of magnitude faster than previous CNN-based methods.

In this paper, we present a real-time salient object detection system based on the minimum spanning tree. Due to the fact that background regions are typically connected to the image boundaries, salient objects can be extracted by computing the distances to the boundaries. However, measuring the image boundary connectivity efficiently is a challenging problem. Existing methods either rely on superpixel representation to reduce the processing units or approximate the distance transform. Instead, we propose an exact and iteration free solution on a minimum spanning tree. The minimum spanning tree representation of an image inherently reveals the object geometry information in a scene. Meanwhile, it largely reduces the search space of shortest paths, resulting an efficient and high quality distance transform algorithm. We further introduce a boundary dissimilarity measure to compliment the shortage of distance transform for salient object detection. Extensive evaluations show that the proposed algorithm achieves the leading performance compared to the state-of-the-art methods in terms of efficiency and accuracy.

The introduction of consumer RGB-D scanners set off a major boost in 3D computer vision research. Yet, the precision of existing depth scanners is not accurate enough to recover fine details of a scanned object. While modern shading based depth refinement methods have been proven to work well with Lambertian objects, they break down in the presence of specularities. We present a novel shape from shading framework that addresses this issue and enhances both diffuse and specular objects' depth profiles. We take advantage of the built-in monochromatic IR projector and IR images of the RGB-D scanners and present a lighting model that accounts for the specular regions in the input image. Using this model, we reconstruct the depth map in real-time. Both quantitative tests and visual evaluations prove that the proposed method produces state of the art depth reconstruction results.