Learning depth from a single image is a challenging task in computer vision. Many recent works on monocular depth estimation explore increasingly large convolutional neural networks to learn monocular cues implicitly. Such methods may fail to generalize well around object boundaries as large networks tend to distort the fine details (such as edges and corners) in low-resolution layers, leading to a

Human-object interaction (HOI) detection is an important vision task that requires the detection of individual object instances and reasoning of their relations. Despite encouraging advancement in recent years, past methods are still limited to relatively simple images where the human and object instances can be detected without difficulties. HOI in the wild should work even when the objects that a

How to extract and integrate spatiotemporal information for video classification is a major challenge. Advanced approaches adopt 2D, and 3D convolution kernels, or their variants as a basis of a spatiotemporal modeling process. However, 2D convolution kernels perform poorly along the temporal dimension, while 3D convolution kernels tend to create confusion between the spatial and temporal sources of

In this paper, we hypothesize that the effects of the degree of typicality in natural semantic categories can be generated based on the structure of artificial categories learned with deep learning models. Motivated by the human approach to representing natural semantic categories and based on the Prototype Theory foundations, we propose a novel Computational Prototype Model (CPM) to represent the

Webly-supervised temporal action localization (WebTAL) leverages web videos to train localization models without requiring manual temporal annotations. WebTAL is extremely challenging since video-level labels on the web are always noisy, seriously damaging the overall performance. Most state-of-the-art methods filter out noise before training, which will inevitably reduce the training samples. In contrast

In recent years, the progress made in deep learning for semantic segmentation has advanced development of semantic visual odometry (VO). Along with point-based and direct methods, VO has recently used edge features. However, mismatches are common in scenes in which the distribution of edges is complex owing to the lack of appropriate descriptors for edges at the present. In this paper, we propose a

Foreground segmentation of moving objects is widely used in different computer vision applications; however, existing deep learning-based methods generally suffer from overall degraded F-measure performance. The two main sources that degrade the F-measure are under-segmentation and catastrophic forgetting. Under-segmentation is the problem of misdetecting objects' fine details. The catastrophic forgetting

Although methods based on the fully convolutional neural networks (FCNs) have shown strong advantages in the field of salient object detection, the existing methods still have two challenging issues: insufficient multi-level feature fusion ability and boundary blur. To overcome these issues, we propose a novel salient object detection method based on a multi-feature fusion cross network (denoted MFC-Net)

Recently, research communities on Computer Vision and biometrics have shown a lot of interest in face verification and classification methods. Fighting against Child Sexual Exploitation Material (CSEM) is one of the applications that might benefit most from these advances. In CSEM, discriminative parts of the face, i.e. mostly the eyes, are often occluded to make the victim identification more difficult

Recent studies use Graph Convolution Networks (GCN) to model label correlation for multi-label images because of the outstanding performance of GCN in relational modeling tasks. However, the traditional GCN has low generalization, and the current state-of-the-arts' accuracy is poor. Therefore, we propose a Two-Stream Dynamic Graph Convolution Network (2S-DGCN) to improve the performance of multi-label

Visual reasoning refers to the process of solving questions about visual information. At present, most visual reasoning models are mainly based on deep learning and end-to-end architecture. Although these models have achieved good performance, they are usually black boxes for users, and it is difficult to understand the basic rationales of the reasoning process. In recent years, the academic community

Long skip connection or encoder-decoder networks for crowd counting have proven to be effective methods to generate high-resolution density maps. However, the simple and coarse feature fusion ignores the disharmony between features, that is, spatial misalignment and semantic inconsistency, which will weaken feature representation and degrade network performance. In this paper, we propose an end-to-end

Domain Adaptation (DA) is a highly relevant research topic when it comes to image classification with deep neural networks. Combining multiple source domains in a sophisticated way to optimize a classification model can improve the generalization to a target domain. Here, the difference in data distributions of source and target image datasets plays a major role. In this paper, we describe based on

Deep Co-Training algorithms are typically comprised of two distinct and diverse feature extractors that simultaneously attempt to learn task-specific features from the same inputs. Achieving such an objective is, however, not trivial, despite its innocent look. This is because homogeneous networks tend to mimic each other under the collaborative training setup. Keeping this difficulty in mind, we make

Recently, intelligent video surveillance applications have become essential in public security by the use of computer vision technologies to investigate and understand long video streams. Anomaly detection and classification are considered a major element of intelligent video surveillance. The aim of anomaly detection is to automatically determine the existence of abnormalities in a short time period

Person re-identification (ReID) is a hot topic in computer vision. The data in the ReID is often collected from the cameras with different views and is affected by other environmental factors, which poses a significant challenge to ReID. The omni-scales proposed by OSNet can extract discriminative feature representations, which shows that omni-scales are adequate for the task of the ReID. However,

Considering that the existing video saliency prediction methods still have limitations in spatiotemporal correlation learning between features and saliency regions, this paper proposes a spatiotemporal module for video saliency prediction based on self-attention. The proposed model emphasizes three essential problems as follows. First, we raise a multi-scale feature-fusion network (MFN) for effective

Nowadays, Sport-related movement recognition plays an essential part in the wellbeing of people's lives. The mention of human movements and gestures is often studied in sports to help analyze, guide, and evaluate activity. The automatic detection of sports-related signals helps find the injuries or indirect physical issues in the human body. Action recognition patterns with complicated motion status

Matrix data has arised in many field, especially in the field of image processing and computer vision. In traditional approaches, the original images need to be vectorized to one-dimension vectors, which may destroy the inherent structure of images. A novel geometrical sparse representation (GSR) model with single image is introduced in this paper that solves a model to measure the similarity between

Classification and regression-based trackers (CAR) are widely adopted to tackle the short-term visual tracking task. However, the existing CAR tackers either employ offline-trained regression models based on predefined anchor-boxes, or online update their models in a rough and inflexible way, which leads to the lack of long-term adaptability for target deformations and appearance variations. To overcome

In the recent years, Convolutional Neural Network-based object detection has experienced impressive progress. Despite these improvements, the performance on the small object detection leaves much to be desired, small objects are often missed or wrongly detected. The reasonable explanation is that the supervisory signals on the small objects are insufficient, and through the analysis of loss distribution

For reality applications of semantic segmentation, inference speed and memory usage are two important factors. To address these challenges, we propose a lightweight feature pyramid encoding network (FPENet) for semantic segmentation with a good trade-off between accuracy and speed. We use a series of feature pyramid encoding (FPE) blocks to encode context at multiple scales in the encoder. Each FPE

Pruning unnecessary regions from weight kernels and filters used in convolutional nerual networks has proven to be effective for model compression. However, prior studies have often over-optimized for generic application areas, which are not ideal for specific problems in salient object detection. In this study, a novel pruning strategy denoted advantage rank (ARank) is developed for multiple salient

Occluded person re-identification (ReID) is a challenging task, which aims at retrieving an occluded person across multiple non-overlapping cameras. To address this issue, we propose a novel framework named Shrinking and Reweighting Network (SRNet) that jointly learns global features by shrinking and reweights part features for matching in an end-to-end framework. Specifically, we use a strong backbone

Unsupervised domain adaptation (UDA) for person re-identication (ReID) remains a challenging task, as the trained ReID system often fails to adapting to a new dataset. Due to the lack of supervision of real labels, the performance of the UDA models suffers from inefficient feature learning and inevitable pseudo label noise. In this work, we tackle the problems by designing an effective dual-path mutual-learning

Facial alignment is an essential task for many higher level facial analysis applications, such as animation, human activity recognition and human - computer interaction. Although the recent availability of big datasets and powerful deep-learning approaches have enabled major improvements on the state of the art accuracy, the performance of current approaches can severely deteriorate when dealing with

In this paper, we consider the problem of domain adaptation for multi-class classification, where we are provided a labeled set of examples in a source dataset and target dataset with no supervision. We tackle the mode collapse problem in adapting the classifier across domains. In this setting, we propose an adversarial learning-based approach using an informative discriminator. Our observation relies

In recent years, discriminative correlation filters (DCF) and Siamese networks based trackers have significantly advanced the performance in tracking. However, the problem of accurate target state estimation is not fully solved yet. Therefore, in this paper, we propose a Criss-Cross Location Attention (CCLA) module, which pays more concerns to global and local contextual information and is used for

Raw point clouds obtained from real-world scanning are always incomplete and ununiformly distributed, which would result in structural losses of object shapes and bring about difficulties in further high-level 3D vision tasks. Therefore, a learning-based method called CRA-Net is proposed in this paper to repair partial point clouds and predict complete object shapes. Compared with most existing networks

Text-based person re-identification (Re-id) is an important task in video surveillance, which consists of retrieving the corresponding person's image given a textual description from a large gallery of images. It is difficult to directly match visual contents with the textual descriptions due to the modality heterogeneity. On the one hand, the textual embedding are not discriminative enough, which

We present a new network architecture able to take advantage of spatio-temporal information available in videos to boost object detection precision. First, box features are associated and aggregated by linking proposals that come from the same anchor box in the nearby frames. Then, we design a new attention module that aggregates short-term enhanced box features to exploit long-term spatio-temporal

This paper presents a domain adaptation strategy to efficiently train network architectures for estimating the relative camera pose in multi-view scenarios. The network architectures are fed by a pair of simultaneously acquired images, hence in order to improve the accuracy of the solutions, and due to the lack of large datasets with pairs of overlapped images, a domain adaptation strategy is proposed

An effective semantic segmentation approach should contain fine object boundaries and continuous regions. However, recent mask-based segmentation cannot extract boundary features well on the coarse prediction, which causes obvious problems of blurry edges. Although several segmentation methods embed boundary detection branch to calculate contours directly, this type of architecture will lead to the

The per-frame segmentation methods have a high computational cost, thereby, these methods are insufficient to cope with the fast inference need of semantic video segmentation. To efficaciously reuse the extracted features by feature propagation, in this paper, we present distortion-aware feature rectification and online selection of keyframes for fast and accurate video segmentation. The proposed dynamic

Fully convolutional network (FCN) based salient object detection methods have shown their advantages in highlighting salient regions because they can obtain global semantic information. And the high-level semantics are usually passed in a top-down pathway. However, the semantic information would be diluted progressively among different level features. To alleviate this issue, we propose a novel edge-aware

Personality computing and affective computing, where the recognition of personality traits is essential, have gained increasing interest and attention in many research areas recently. We propose a novel approach to recognize the Big Five personality traits of people from videos. To this end, we use four different modalities, namely, ambient appearance (scene), facial appearance, voice, and transcribed

Visual Question Answering (VQA) has recently appeared as a hot research area in the field of computer vision and natural language processing. A VQA model uses both image and question features and fuses them to predict an answer for a given natural question related to an image. However, most VQA approaches using attention mechanism mainly concentrate on extraction of visual information from regions

Learning from limited supervision is a challenging problem that has recently attracted wide attention in the machine learning community. With scarce annotated samples available in target categories, so-called few-shot image recognition aims to transfer basic knowledge from a large-scale image set to recognize unseen classes. Many existing approaches tend to learn a general source data representation

A power substation is susceptible to intrusions of foreign objects. The intrusions can likely result in failures of power supplies. Therefore, recognizing foreign objects becomes important to ensure constant and stable power supplies. However, existing object recognition methods fail to achieve acceptable accuracy and performance. In this paper, we propose an efficient Foreign Objects Detection Network

Semantic segmentation and instance segmentation based on 3D point clouds involve significant challenges, specifically in the task of joint semantic and instance segmentation. The efficient and effective mutual assistance between semantic and instance segmentation is rarely considered and still remains an unaddressed research problem. To address this, herein, a novel and robust 3D point cloud segmentation

Network convergence as well as recognition accuracy are essential issues when applying Convolutional Neural Networks (CNN) to human action recognition. Most deep learning methods neglect model convergence when striving to improve the abstraction capability, thus degrading the performances sharply when computing resources are limited. To mitigate this problem, we propose a structure named 2D Progressive

We propose an approach for quickly building a visual representation of a full indoor building. Our goal is to enable intelligent systems which frequently and regularly monitor buildings to assist personnel operating remotely, a need of special importance in these days. Prior work in neural scene representations for view synthesis focuses on single objects and small scenes and does not scale to full

Efficient semantic segmentation of large-scale three-dimensional (3D) point clouds is an essential approach for intelligent robots to perceive the surrounding environment. However, due to the expensive sampling process or time-consuming pre/post-processing steps, most of the current solutions are inefficient or limited in scale. In this paper, we propose a novel framework, ThickSeg, to efficiently

In this paper, we propose a novel framework to segment 3D human motion capture data into distinct behaviors. First, in preprocessing, we build a normalized pose space by eliminating translation and orientation from the 3D poses. We then transform these normalized 3D poses into 2D RGB images, and as a result, we simplify the task of motion segmentation as image classification and recognition. Furthermore

Visual attention mechanism has been widely used by image captioning model in order to dynamically attend to the related visual region based on given language information. Such capability allows a trained model to carry out fine-grained level image understanding and reasoning. However, existing visual attention models only focus on the individual visual region in the image and the alignment between

This paper proposes a deep convolutional neural network with a concise and effective encoder-decoder architecture for saliency prediction. Local and global contextual features make a considerable contribution to saliency prediction. In order to integrate and exploit these features more thoroughly, in the proposed pithy architecture, we deploy a dense and global context connection structure between

The model pruning is one of the predominant model compression tasks to decrease the demands in computing power and memory footprint. However, most existing pruning methods have overly broad application areas, which defects in a sub-optimal solution specifically to solve certain specified difficult problems in the tasks of salient object detection. In this paper, we propose a novel solution, dubbed

Recently, many efforts have been made to model spatial–temporal features from human skeleton for action recognition by using graph convolutional networks (GCN). Skeleton sequence can precisely represent human pose with a small number of joints while there is still a lot of redundancies across the skeleton sequence in the term of temporal dependency. In order to improve the effectiveness of spatial–temporal