Image segmentation is a key topic in image processing and computer vision with applications such as scene understanding, medical image analysis, robotic perception, video surveillance, augmented reality, and image compression, among many others. Various algorithms for image segmentation have been developed in the literature. Recently, due to the success of deep learning models in a wide range of vision applications, there has been a substantial amount of works aimed at developing image segmentation approaches using deep learning models. In this survey, we provide a comprehensive review of the literature at the time of this writing, covering a broad spectrum of pioneering works for semantic and instance-level segmentation, including fully convolutional pixel-labeling networks, encoder-decoder architectures, multi-scale and pyramid based approaches, recurrent networks, visual attention models, and generative models in adversarial settings. We investigate the similarity, strengths and challenges of these deep learning models, examine the most widely used datasets, report performances, and discuss promising future research directions in this area.

Image Segmentation Using Deep Learning: A Survey

Improving person re-identification by attribute and identity learning

Person re-identification (Re-ID) aims at retrieving a person of interest across multiple non-overlapping cameras. With the advancement of deep neural networks and increasing demand of intelligent video surveillance, it has gained significantly increased interest in the computer vision community. By dissecting the involved components in developing a person Re-ID system, we categorize it into the closed-world and open-world settings. The widely studied closed-world setting is usually applied under various research-oriented assumptions, and has achieved inspiring success using deep learning techniques on a number of datasets. We first conduct a comprehensive overview with in-depth analysis for closed-world person Re-ID from three different perspectives, including deep feature representation learning, deep metric learning and ranking optimization. With the performance saturation under closed-world setting, the research focus for person Re-ID has recently shifted to the open-world setting, facing more challenging issues. This setting is closer to practical applications under specific scenarios. We summarize the open-world Re-ID in terms of five different aspects. By analyzing the advantages of existing methods, we design a powerful AGW baseline, achieving state-of-the-art or at least comparable performance on twelve datasets for FOUR different Re-ID tasks. Meanwhile, we introduce a new evaluation metric (mINP) for person Re-ID, indicating the cost for finding all the correct matches, which provides an additional criteria to evaluate the Re-ID system for real applications. Finally, some important yet under-investigated open issues are discussed.

Deep Learning for Person Re-identification: A Survey and Outlook

Learning Deconvolution Network for Semantic Segmentation

Person re-identification (re-ID) is mostly viewed as an image retrieval problem. This task aims to search a query person in a large image pool. In practice, person re-ID usually adopts automatic detectors to obtain cropped pedestrian images. However, this process suffers from two types of detector errors: excessive background and part missing. Both errors deteriorate the quality of pedestrian alignment and may compromise pedestrian matching due to the position and scale variances. To address the misalignment problem, we propose that alignment be learned from an identification procedure. We introduce the pedestrian alignment network (PAN) which allows discriminative embedding learning pedestrian alignment without extra annotations. We observe that when the convolutional neural network learns to discriminate between different identities, the learned feature maps usually exhibit strong activations on the human body rather than the background. The proposed network thus takes advantage of this attention mechanism to adaptively locate and align pedestrians within a bounding box. Visual examples show that pedestrians are better aligned with PAN. Experiments on three large-scale re-ID datasets confirm that PAN improves the discriminative ability of the feature embeddings and yields competitive accuracy with the state-of-the-art methods.

Pedestrian Alignment Network for Large-scale Person Re-Identification

Partial person re-identification (re-id) is a challenging problem, where only several partial observations (images) of people are available for matching. However, few studies have provided flexible solutions to identifying a person in an image containing arbitrary part of the body. In this paper, we propose a fast and accurate matching method to address this problem. The proposed method leverages Fully Convolutional Network (FCN) to generate fix-sized spatial feature maps such that pixel-level features are consistent. To match a pair of person images of different sizes, a novel method called Deep Spatial feature Reconstruction (DSR) is further developed to avoid explicit alignment. Specifically, DSR exploits the reconstructing error from popular dictionary learning models to calculate the similarity between different spatial feature maps. In that way, we expect that the proposed FCN can decrease the similarity of coupled images from different persons and increase that from the same person. Experimental results on two partial person datasets demonstrate the efficiency and effectiveness of the proposed method in comparison with several state-of-the-art partial person re-id approaches. Additionally, DSR achieves competitive results on a benchmark person dataset Market1501 with 83.58% Rank-1 accuracy.

/pdf/deep-spatial-feature-reconstruction-for-partial-person-re-4pzfmk3c5s.pdf

Deep Spatial Feature Reconstruction for Partial Person Re-identification: Alignment-free Approach

Conventional iris recognition requires controlled conditions (e.g., close acquisition distance and stop-and-stare scheme) and high user cooperation for image acquisition. Non-cooperative acquisition environments introduce many adverse factors such as blur, off-axis, occlusions and specular reflections, which challenge existing iris segmentation approaches. In this paper, we present two iris segmentation models, namely hierarchical convolutional neural networks (HCNNs) and multi-scale fully convolutional network (MFCNs), for noisy iris images acquired at-a-distance and on-the-move. Both models automatically locate iris pixels without handcrafted features or rules. Moreover, the features and classifiers are jointly optimized. They are end-to-end models which require no further pre- and post-processing and outperform other state-of-the-art methods. Compared with HCNNs, MFCNs take input of arbitrary size and produces correspondingly-sized output without sliding window prediction, which makes MFCNs more efficient. The shallow, fine layers and deep, global layers are combined in MFCNs to capture both the texture details and global structure of iris patterns. Experimental results show that MFCNs are more robust than HCNNs to noises, and can greatly improve the current state-of-the-arts by 25.62% and 13.24% on the UBIRIS.v2 and CASIA.v4-distance databases, respectively.

Accurate iris segmentation in non-cooperative environments using fully convolutional networks

DeepIris: Learning Pairwise Filter Bank for Heterogeneous Iris Verification

The quality of iris images on mobile devices is significantly degraded due to hardware limitations and less constrained environments. Traditional iris recognition methods cannot achieve high identification rate using these low-quality images. To enhance the performance of mobile identification, we develop a deep feature fusion network that exploits the complementary information presented in iris and periocular regions. The proposed method first applies maxout units into the convolutional neural networks (CNNs) to generate a compact representation for each modality and then fuses the discriminative features of two modalities through a weighted concatenation. The parameters of convolutional filters and fusion weights are simultaneously learned to optimize the joint representation of iris and periocular biometrics. To promote the iris recognition research on mobile devices under near-infrared (NIR) illumination, we publicly release the CASIA-Iris-Mobile-V1.0 database, which in total includes 11 000 NIR iris images of both eyes from 630 Asians. It is the largest NIR mobile iris database as far as we know. On the newly built CASIA-Iris-M1-S3 data set, the proposed method achieves 0.60% equal error rate and 2.32% false non-match rate at false match rate $=10^{-5}$ , which are obviously better than unimodal biometrics as well as traditional fusion methods. Moreover, the proposed model requires much fewer storage spaces and computational resources than general CNNs.

Deep Feature Fusion for Iris and Periocular Biometrics on Mobile Devices

Partial person re-identification (re-id) is a challenging problem, where only several partial observations (images) of people are available for matching. However, few studies have provided flexible solutions to identifying a person in an image containing arbitrary part of the body. In this paper, we propose a fast and accurate matching method to address this problem. The proposed method leverages Fully Convolutional Network (FCN) to generate fix-sized spatial feature maps such that pixel-level features are consistent. To match a pair of person images of different sizes, a novel method called Deep Spatial feature Reconstruction (DSR) is further developed to avoid explicit alignment. Specifically, DSR exploits the reconstructing error from popular dictionary learning models to calculate the similarity between different spatial feature maps. In that way, we expect that the proposed FCN can decrease the similarity of coupled images from different persons and increase that from the same person. Experimental results on two partial person datasets demonstrate the efficiency and effectiveness of the proposed method in comparison with several state-of-the-art partial person re-id approaches. Additionally, DSR achieves competitive results on a benchmark person dataset Market1501 with 83.58\% Rank-1 accuracy.

Haiqing Li

Papers

Deep Spatial Feature Reconstruction for Partial Person Re-identification: Alignment-free Approach

Accurate iris segmentation in non-cooperative environments using fully convolutional networks

DeepIris: Learning Pairwise Filter Bank for Heterogeneous Iris Verification

Deep Feature Fusion for Iris and Periocular Biometrics on Mobile Devices

Deep Spatial Feature Reconstruction for Partial Person Re-identification: Alignment-Free Approach