Semantic labeling of RGB-D scenes is crucial to many intelligent applications including perceptual robotics. It generates pixelwise and fine-grained label maps from simultaneously sensed photometric (RGB) and depth channels. This paper addresses this problem by i) developing a novel Long Short-Term Memorized Context Fusion (LSTM-CF) Model that captures and fuses contextual information from multiple channels of photometric and depth data, and ii) incorporating this model into deep convolutional neural networks (CNNs) for end-to-end training. Specifically, contexts in photometric and depth channels are, respectively, captured by stacking several convolutional layers and a long short-term memory layer; the memory layer encodes both short-range and long-range spatial dependencies in an image along the vertical direction. Another long short-term memorized fusion layer is set up to integrate the contexts along the vertical direction from different channels, and perform bi-directional propagation of the fused vertical contexts along the horizontal direction to obtain true 2D global contexts. At last, the fused contextual representation is concatenated with the convolutional features extracted from the photometric channels in order to improve the accuracy of fine-scale semantic labeling. Our proposed model has set a new state of the art, i.e., 48.1% and 49.4% average class accuracy over 37 categories (2.2% and 5.4% improvement) on the large-scale SUNRGBD dataset and the NYUDv2dataset, respectively.
Recently, machine learning based single image super resolution (SR) approaches focus on jointly learning representations for high-resolution (HR) and low-resolution (LR) image patch pairs to improve the quality of the super-resolved images. However, due to treat all image pixels equally without considering the salient structures, these approaches usually fail to produce visual pleasant images with sharp edges and fine details. To address this issue, in this work we present a new novel SR approach, which replaces the main building blocks of the classical interpolation pipeline by a flexible, content-adaptive deep neural networks. In particular, two well-designed structure-aware components, respectively capturing local- and holistic- image contents, are naturally incorporated into the fully-convolutional representation learning to enhance the image sharpness and naturalness. Extensively evaluations on several standard benchmarks (e.g., Set5, Set14 and BSD200) demonstrate that our approach can achieve superior results, especially on the image with salient structures, over many existing state-of-the-art SR methods under both quantitative and qualitative measures.
Cross-domain visual data matching is one of the fundamental problems in many real-world vision tasks, e.g., matching persons across ID photos and surveillance videos. Conventional approaches to this problem usually involves two steps: i) projecting samples from different domains into a common space, and ii) computing (dis-)similarity in this space based on a certain distance. In this paper, we present a novel pairwise similarity measure that advances existing models by i) expanding traditional linear projections into affine transformations and ii) fusing affine Mahalanobis distance and Cosine similarity by a data-driven combination. Moreover, we unify our similarity measure with feature representation learning via deep convolutional neural networks. Specifically, we incorporate the similarity measure matrix into the deep architecture, enabling an end-to-end way of model optimization. We extensively evaluate our generalized similarity model in several challenging cross-domain matching tasks: person re-identification under different views and face verification over different modalities (i.e., faces from still images and videos, older and younger faces, and sketch and photo portraits). The experimental results demonstrate superior performance of our model over other state-of-the-art methods.
The past decade has witnessed the rapid development of feature representation learning and distance metric learning, whereas the two steps are often discussed separately. To explore their interaction, this work proposes an end-to-end learning framework called DARI, i.e. Distance metric And Representation Integration, and validates the effectiveness of DARI in the challenging task of person verification. Given the training images annotated with the labels, we first produce a large number of triplet units, and each one contains three images, i.e. one person and the matched/mismatch references. For each triplet unit, the distance disparity between the matched pair and the mismatched pair tends to be maximized. We solve this objective by building a deep architecture of convolutional neural networks. In particular, the Mahalanobis distance matrix is naturally factorized as one top fully-connected layer that is seamlessly integrated with other bottom layers representing the image feature. The image feature and the distance metric can be thus simultaneously optimized via the one-shot backward propagation. On several public datasets, DARI shows very promising performance on re-identifying individuals cross cameras against various challenges, and outperforms other state-of-the-art approaches.
This paper investigates how to rapidly and accurately localize facial landmarks in unconstrained, cluttered environments rather than in the well segmented face images. We present a novel Backbone-Branches Fully-Convolutional Neural Network (BB-FCN), which produces facial landmark response maps directly from raw images without relying on pre-process or sliding window approaches. BB-FCN contains one backbone and a number of network branches with each corresponding to one landmark type, and it operates in a progressive manner. Specifically, the backbone roughly detects the locations of facial landmarks by taking the whole image as input, and the branches further refine the localizations based on a local observation from the backbone's intermediate feature map. Moreover, our backbone-branches architecture does not contain full-connection layers for location regression, leading to efficient learning and inference. Our extensive experiments show that our model achieves superior performances over other state-of-the-arts under both the constrained (i.e. with face regions) and the "in the wild" scenarios.
This paper addresses the problem of geometric scene parsing, i.e. simultaneously labeling geometric surfaces (e.g. sky, ground and vertical plane) and determining the interaction relations (e.g. layering, supporting, siding and affinity) between main regions. This problem is more challenging than the traditional semantic scene labeling, as recovering geometric structures necessarily requires the rich and diverse contextual information. To achieve these goals, we propose a novel recurrent neural network model, named Hierarchical Long Short-Term Memory (H-LSTM). It contains two coupled sub-networks: the Pixel LSTM (P-LSTM) and the Multi-scale Super-pixel LSTM (MS-LSTM) for handling the surface labeling and relation prediction, respectively. The two sub-networks provide complementary information to each other to exploit hierarchical scene contexts, and they are jointly optimized for boosting the performance. Our extensive experiments show that our model is capable of parsing scene geometric structures and outperforming several state-of-the-art methods by large margins. In addition, we show promising 3D reconstruction results from the still images based on the geometric parsing.
By taking the semantic object parsing task as an exemplar application scenario, we propose the Graph Long Short-Term Memory (Graph LSTM) network, which is the generalization of LSTM from sequential data or multi-dimensional data to general graph-structured data. Particularly, instead of evenly and fixedly dividing an image to pixels or patches in existing multi-dimensional LSTM structures (e.g., Row, Grid and Diagonal LSTMs), we take each arbitrary-shaped superpixel as a semantically consistent node, and adaptively construct an undirected graph for each image, where the spatial relations of the superpixels are naturally used as edges. Constructed on such an adaptive graph topology, the Graph LSTM is more naturally aligned with the visual patterns in the image (e.g., object boundaries or appearance similarities) and provides a more economical information propagation route. Furthermore, for each optimization step over Graph LSTM, we propose to use a confidence-driven scheme to update the hidden and memory states of nodes progressively till all nodes are updated. In addition, for each node, the forgets gates are adaptively learned to capture different degrees of semantic correlation with neighboring nodes. Comprehensive evaluations on four diverse semantic object parsing datasets well demonstrate the significant superiority of our Graph LSTM over other state-of-the-art solutions.
Maximally stable extremal regions (MSER), which is a popular method to generate character proposals/candidates, has shown superior performance in scene text detection. However, the pixel-level operation limits its capability for handling some challenging cases (e.g., multiple connected characters, separated parts of one character and non-uniform illumination). To better tackle these cases, we design a character proposal network (CPN) by taking advantage of the high capacity and fast computing of fully convolutional network (FCN). Specifically, the network simultaneously predicts characterness scores and refines the corresponding locations. The characterness scores can be used for proposal ranking to reject non-character proposals and the refining process aims to obtain the more accurate locations. Furthermore, considering the situation that different characters have different aspect ratios, we propose a multi-template strategy, designing a refiner for each aspect ratio. The extensive experiments indicate our method achieves recall rates of 93.88%, 93.60% and 96.46% on ICDAR 2013, SVT and Chinese2k datasets respectively using less than 1000 proposals, demonstrating promising performance of our character proposal network.
Sampling and budgeting training examples are two essential factors in tracking algorithms based on support vector machines (SVMs) as a trade-off between accuracy and efficiency. Recently, the circulant matrix formed by dense sampling of translated image patches has been utilized in correlation filters for fast tracking. In this paper, we derive an equivalent formulation of a SVM model with circulant matrix expression and present an efficient alternating optimization method for visual tracking. We incorporate the discrete Fourier transform with the proposed alternating optimization process, and pose the tracking problem as an iterative learning of support correlation filters (SCFs) which find the global optimal solution with real-time performance. For a given circulant data matrix with n^2 samples of size n*n, the computational complexity of the proposed algorithm is O(n^2*logn) whereas that of the standard SVM-based approaches is at least O(n^4). In addition, we extend the SCF-based tracking algorithm with multi-channel features, kernel functions, and scale-adaptive approaches to further improve the tracking performance. Experimental results on a large benchmark dataset show that the proposed SCF-based algorithms perform favorably against the state-of-the-art tracking methods in terms of accuracy and speed.
Identifying the same individual across different scenes is an important yet difficult task in intelligent video surveillance. Its main difficulty lies in how to preserve similarity of the same person against large appearance and structure variation while discriminating different individuals. In this paper, we present a scalable distance driven feature learning framework based on the deep neural network for person re-identification, and demonstrate its effectiveness to handle the existing challenges. Specifically, given the training images with the class labels (person IDs), we first produce a large number of triplet units, each of which contains three images, i.e. one person with a matched reference and a mismatched reference. Treating the units as the input, we build the convolutional neural network to generate the layered representations, and follow with the $L2$ distance metric. By means of parameter optimization, our framework tends to maximize the relative distance between the matched pair and the mismatched pair for each triplet unit. Moreover, a nontrivial issue arising with the framework is that the triplet organization cubically enlarges the number of training triplets, as one image can be involved into several triplet units. To overcome this problem, we develop an effective triplet generation scheme and an optimized gradient descent algorithm, making the computational load mainly depends on the number of original images instead of the number of triplets. On several challenging databases, our approach achieves very promising results and outperforms other state-of-the-art approaches.