Bailin Yang

• Doctor of Philosophy
• Zhejiang Gongshang University

In the increasingly digitized world, the privacy and security of sensitive data shared via IoT devices are paramount. Traditional privacy-preserving methods like k-anonymity and ldiversity are becoming outdated due to technological advancements. In addition, data owners often worry about misuse and unauthorized access to their personal information....

The advancement of new technology with an adaptation of the Internet of Things (IoTs) has created a new digital-commerce paradigm. IoT service accessibility offers numerous benefits, but it also puts customer-related data at risk. A novel authentication method is presented for the linked devices in the IoTs to meet security problems and enable ligh...

We present a novel framework that can generate plausible and diverse 3D (3 Dimensions) indoor scenes based on room floor plans and user-specified layout objects. In the framework, we construct a generative neural network with a non-autoregressive transformer to generate a reasonable distribution of layout objects, and then apply a fine-grained opti...

Self-supervised monocular depth estimation has opened up exciting possibilities for practical applications, including scene understanding, object detection, and autonomous driving, without the need for expensive depth annotations. However, traditional methods for single-image depth estimation encounter limitations in photometric loss due to a lack...

Scene cartoonization aims to convert photos into stylized cartoons. While GANs can generate high-quality images, previous methods focus on individual images or single styles, ignoring relationships between datasets. We propose a novel multi-style scene cartoonization GAN that leverages multiple cartoon datasets jointly. Our main technical contribut...

Color composition (or color theme) is a key factor to determine how well a piece of art work or graphical design is perceived by humans. Despite a few color harmony models have been proposed, their results are often less satisfactory since they mostly neglect the variations of aesthetic cognition among individuals and treat the influence of all rat...

Scene cartoonization aims to convert photos into stylized cartoons. While generative adversarial networks (GANs) can generate high‐quality images, previous methods focus on individual images or single styles, ignoring relationships between datasets. We propose a novel multi‐style scene cartoonization GAN that leverages multiple cartoon datasets joi...

We present a novel framework that can generate plausible and diverse 3D (3 Dimensions) indoor scenes based on room floor plans and user-specified layout objects. In the framework, we construct a generative neural network with a non-autoregressive transformer to generate a reasonable distribution of layout objects, and then apply a fine-grained opti...

We present a novel approach for modeling artists’ drawing processes using an architecture that combines an unconditional generative adversarial network (GAN) with a multi-view generator and multi-discriminator. Our method excels in synthesizing various types of picture drawing, including line drawing, shading, and color drawing, achieving high qual...

We propose an automatic layout method for indoor scenes that effectively satisfies specific constraints. Our approach involves enhancing the existing scene representation method to accommodate complex constraints, including the precise placement of doors, windows, and user-specified furniture. To achieve this, we construct a conditional vector that...

Ensuring public safety in urban areas is a crucial element in maintaining a good quality of life. The successful deployment of video surveillance systems depends heavily on the acquisition and processing of large volumes of urban data to derive meaningful insights. Manual monitoring and analysis of anomalous activities in the surveillance footage i...

Deinterlacing is a classical issue in video processing, aimed at generating progressive video from interlaced content. There are precious videos that are difficult to reshoot and still contain interlaced content. Previous methods have primarily focused on simple interlaced mechanisms and have struggled to handle the complex artifacts present in rea...

We present a novel approach to differentiable rendering for participating media, addressing the challenge of computing scene parameter derivatives. While existing methods focus on derivative computation within volumetric path tracing, they fail to significantly improve computational performance due to the expensive computation of multiply‐scattered...

Information-theoretic measures have been commonly applied to evaluate the relevance and redundancy in multi-label feature selection. However, the current multi-label feature selection methods based on information-theoretic measures neglect the dynamic changes in the relevance of selected features and candidate features. Furthermore, they also do no...

The concept of 2 Dimensional (2D) and 3 Dimensional (3D) are represented by different bodies for diverse applications. The 2D is an old concept of image representation as it displays only the x and y axis, whereas the 3D image displays the x, y, and z axis simultaneously. The image of 3D on a screen looks like an image in the real world. With the r...

The image quality assessment (IQA) for medical images has been challenging due to their usage for diagnostic purposes. Traditional convolutional-neural-network-based IQA models usually assess the image quality on a global scale, which is unsuitable for inferring the local quality of medical images from the diagnostic attention perspective. To allev...

Knowledge graph representation learning aims to embed kn-owledge facts into a continuous vector space, enabling models to capture semantic connections within and between triples. However, existing methods primarily focus on a single dimension of entities or relations, limiting their ability to learn knowledge facts. To address this issue, this pape...

Judging how an image is visually appealing is a complicated and subjective task. This highly motivates having a machine learning model to automatically evaluate image aesthetic by matching the aesthetics of general public. Although deep learning methods have been successfully learning good visual features from images, correctly assessing image aest...

Reconstructing a 3D face from a single image is a crucial task in numerous multimedia applications. Face images with ground-truth 3D face shapes are scarce, so unsupervised deep learning methods, which rely primarily on the free supervision signal derived from the visual disparity between the input image and the rendered counterpart of the predicte...

The amalgamation of Smart IoT and Machine learning is an emerging research area. In this context, a new trend in IoT called Social IoT has been considered for this study. The Social IoT has benefits of connectivity exhibited within the network of connected objects through the Internet of Things (IoT). It covers the entire world and provides innovat...

Recently, computer-aided diagnosis (CAD) systems powered by deep learning (DL) algorithms have shown excellent performance in the evaluation of digital mammography for breast cancer diagnosis. However, such systems typically require pixel-level annotations by expert radiologists which is prohibitively time-consuming and expensive. Medical institute...

Retinal vessels play an important role in judging many eye-related diseases, so accurate segmentation of retinal vessels has become the key to auxiliary diagnosis. In this paper, we present a Cascaded Residual Attention U-Net(CRAUNet) that can be regarded as a set of U-Nets, that allows coarse-to-fine representations. In the CRAUNet, we introduce a...

Real-time violence detection with the use of surveillance is the process of using live videos to detect violent and irregular behavior. In organizations, they use some potential procedures for recognition the activity in which normal and abnormal activities can be found easily. In this research, multiple key challenges have been oncorporated with t...

Objectives: Develop and evaluate the performance of deep learning and linear regression cascade algorithms for automated assessment of the image layout and position of chest radiographs. Methods: This retrospective study used 10 quantitative indices to capture subjective perceptions of radiologists regarding image layout and position of chest ra...

Purpose Due to the incompleteness nature of knowledge graphs (KGs), the task of predicting missing links between entities becomes important. Many previous approaches are static, this posed a notable problem that all meanings of a polysemous entity share one embedding vector. This study aims to propose a polysemous embedding approach, named KG embed...

Underwater Wireless Sensor Network (UWSN) accomplishes the consideration of a few scientists and academicians towards itself. Because of the brutality of the climate lies submerged represents various difficulties, i.e., high transmission delay, outstanding piece mistake rate, more expense in usage, sinks development and energy imperatives, unequal...

The depth image based rendering (DIBR) is a key technology in the emerging 3D video and 3D-TV, which uses a single texture image and the corresponding depth map to generate the virtual views. A critical problem occurs when the virtual view generation refers to regions covered by foreground objects in the original view, which might be disoccluded in...

Multi-frame human pose estimation in complicated situations is challenging. Although state-of-the-art human joints detectors have demonstrated remarkable results for static images, their performances come short when we apply these models to video sequences. Prevalent shortcomings include the failure to handle motion blur, video defocus, or pose occ...

In this paper, we present a segmentation and classification method for thyroid follicular neoplasm based on the combination of the prior-based level set method and deep convolutional neural network (DCNN).The proposed method aims to discriminate thyroid follicular adenoma (TFA) and follicular thyroid carcinoma (FTC) in ultrasound images. From their...

A new automatic layout scheme for interior furniture is presented. According to user-specified furniture, an empty room region is divided into several functional areas by use of conditional generative adversarial networks. Each type of functional area can contain some categories of particular furniture objects, while each category of furniture belo...

The elastic network models (ENMs) are known as representative coarse-grained models to capture essential dynamics of proteins. Due to simple designs of the force constants as a decay with spatial distances of residue pairs in many previous studies, there is still much room for the improvement of ENMs. In this article, we directly computed the force...

To effectively control particle‐based cloud evolution without imposing strict position constraints, we propose a novel method integrating a control force field and a phase transition control into the position‐based fluids (PBF) framework. To produce realistic cloud simulation, we incorporate both fluid dynamics and thermodynamics to govern cloud pa...

Knowledge graphs (KGs) have achieved great success in many AI-related applications in the past decade. Although KGs contain billions of real facts, they are usually not complete. This problem arises to the task of missing link prediction whose purpose is to perform link prediction between entities. Knowledge graph embedding has proved to be a highl...

Physically-based cloud simulation is an effective approach for synthesizing realistic cloud. However, generating clouds with desired shapes requires a time-consuming process for selecting the appropriate simulation parameters. This paper addresses such a problem by solving an inverse cloud forming problem. We propose a convolutional neural network,...

In vehicle retrieval, the vehicle patch should first be localized to remove the irrelevant background information. Moreover, the negative samples are much more prevalent than the positive samples, and the information from the negative samples is not fully exploited in the triple loss. What we need is a way to incorporate global knowledge and struct...

The segmentation of video, or separating out objects in the foreground, is an important application of pattern recognition and computer vision. Segmentation errors in pattern recognition approaches mainly come from difficulties in selecting maximally informative frames for learning. In this paper, we develop an approach to video segmentation that r...

The plots of many literary works are very complex, which hinders the readers' comprehension of these literary works. Thus, to help readers' comprehension of complex literary works, tools should be proposed to support their comprehension by presenting the most important information to readers. In the case of literary works, the most important inform...

Color structure of a home scene image closely relates to the material properties of its local regions. Existing color migration methods typically fail to fully infer the correlation between the coloring of local home scene regions, leading to a local blur problem. In this paper, we propose a color migration framework for home scene images. It picks...

Harmonious color combinations can stimulate positive user emotional responses. However, a widely open research question is: how can we establish a robust and accurate color harmony measure for the public and professional designers to identify the harmony level of a color theme or color set. Building upon the key discovery that color pairs play an i...

Writing is an important basic skill for humans. To acquire such a skill, pupils often have to practice writing for several hours each day. However, different pupils usually possess distinct writing postures. Bad postures not only affect the speed and quality of writing, but also severely harm the healthy development of pupils’ spine and eyesight. T...

Certificateless strong designated verifier signature schemes have realized the merit of CL-PKC against the traditional strong designated verifier signatures. However, when the signer and the designated verifier disagree with the signature, existing schemes cannot distinguish the original signature from the signature transcript. In addition, errors...

Internet of things (IoT) can enable cyber-physical objects to communicate with one another and realize real-time living-needs control such as for vehicles, smart phones, refrigerators, healthcare gadgets and air-conditioners. Of the applications of IoT, collecting and receiving health-related data securely is the most crucial and significant use in...

Depth-image-based rendering (DIBR) is widely used in 3DTV, free-viewpoint video, and interactive 3D graphics applications. Typically, synthetic images generated by DIBR-based systems incorporate various distortions, particularly geometric distortions induced by object dis-occlusion. Ensuring the quality of synthetic images is critical to maintainin...

Recent research on dense captioning based on the recurrent neural network and the convolutional neural network has made a great progress. However, mapping from an image feature space to a description space is a nonlinear and multimodel task, which makes it difficult for the current methods to get accurate results. In this paper, we put forward a no...

With the increasing demand in using 3D mesh data over networks, supporting effective compression and efficient transmission of meshes has caught lots of attention in recent years. This article introduces a novel compression method for 3D mesh animation sequences, supporting user-defined and progressive transmissions over networks. Our motion-aware...

Indoor home scene coloring technology is a hot topic for home design, helping users make home coloring decisions. Image based home scene coloring is preferable for e-commerce customers since it only requires users to describe coloring expectations or manipulate colors through images, which is intuitive and inexpensive. In contrast, if home scene co...

Realistic cloud is essential for enhancing the quality of computer graphics applications, such as flight simulation. Data-driven method is an effective way in cloud modeling, but existing methods typically only utilize one data source as input. For example, natural images are usually used to model small-scale cloud with details, and satellite image...

Traffic flow prediction plays a key role in intelligent transportation systems. However, since traffic sensors are typically manually controlled, traffic flow data with varying length, irregular sampling and missing data are difficult to exploit effectively. To overcome this problem, we propose a novel approach that is based on Long Short-Term Memo...

Depth-image-based rendering (DIBR) is widely used to support 3D interactive graphics on low-end mobile devices. Although it reduces the rendering cost on a mobile device, it essentially turns such a cost into depth image transmission cost or bandwidth consumption, inducing performance bottleneck to a remote rendering system. To address this problem...

Dynamic mesh sequence (DMS) is a simple and accurate representation for precisely recording a 3D animation sequence. Despite its simplicity, this representation is typically large in data size, making storage and transmission expensive. This paper presents a novel framework that allows effective DMS compression and progressive streaming by eliminat...

By collecting the data of eyeball movement of pilots, it is possible to monitor pilot's operation in the future flight in order to detect potential accidents. In this paper, we designed a novel SVS system that is integrated with an eye tracking device, and is able to achieve the following functions:1) A novel method that is able to learn from the e...

Image segmentation is a necessary but difficult task in many image processing applications. Unlike conventional auto-segmentation, semi-supervised image segmentation involves a moderate amount of user interaction. In this paper, a novel interactive image segmentation method based on diffusion maps is proposed, which can better account for the distr...

This paper proposes an encoding and reconstructing method with robust transmission for 3D model topological data. For the encoding of topological data, we firstly adopt the valence-driven method to visit the full mesh and give each vertex a sequence number, degree information and the neighbouring vertex information. Then, we adopt the improved grap...

Mobile robots are useful for environment exploration and rescue operations. In such applications, it is crucial to accurately analyse and represent an environment, providing appropriate inputs for motion planning in order to support robot navigation and operations. 2D mapping methods are simple but cannot handle multilevel or multistory environment...

Reaching is one of the most important behaviors in our daily life and has attracted plenty of researchers to work on it both in computer animation and robot research area. However, existing proposed methods either lack of flexibility or their results are not so convincing. In this paper, we present a novel controller-based framework for reaching mo...

Deformation method based on moving least squares (MLS) allows the user to manipulate 2D characters using either sets of points or line segments in real time. However, the traditional MLS deformation spreads the deformation of the controls with respect to the spatial distance, but oblivious to the shape topology, which would possibly lead to distort...

In this paper, we present a novel variational model for salt and pepper noise removal, and an efficient numerical algorithm for solving it. The proposed model features the use of an approximating function of norm to measure the closeness of the reconstructed and observed images at the pixels which are not the candidates of the noisy pixels. In addi...

Mesh geometry can be used to model both object shape and details. If texture maps are involved, it is common to let mesh geometry mainly model object shapes and let the texture maps model the most object details, optimising data size and complexity of an object. To support efficient object rendering and transmission, model simplification can be app...

Different from common 2D images, when a texture map image is project to a 3D model in the 3D space, it also implicitly associates with certain 3D geometry information. However, existing common texture map image compression methods do not take this into account. In this paper, we present a visual importance driven selective compression method for te...

Realtime streaming 3D model over wireless network is a challenging proposition due to the characteristic of 3D progressive geometry data and wireless network. In this paper, we propose a perceptual-driven 3D progressive model selective transmission method. For the progressive model, the data including the base mesh and a series of vertex splits wil...

Differing from common 2D images, a texture map, since it is used to project onto a 3D model in 3D space, not only contains 2D texture information, but also implicitly associates certain 3D geometric information. Related to this, an effective 3D geometry-dependent texture map compression method with hybrid region of interest (ROI) coding is proposed...

Multiple versions will be produced in the process of complex products developing. There are mass data of product components in different types and formats and relationships between components. Traditional researches of version management model concentrate on the information of single object rather than that of multi-objects and relationships betwee...

As we know, the mobile device screen is small. The lower accuracy of the model is relatively weak and the capacity to handle high detailed model is very limited. What's more, the existing three-dimensional simplification algorithms are for the personal computer and they are not suitable for the mobile terminals. Thus, we propose a novel 3D model si...

In this paper, we present a visual important driven selective compression method for texture map image. It is based on the notion that the visual important area of texture map image is regarded as texel region of interest, which will be compressed less than the other background areas. In order to obtain the visual important areas, we take not only...

This paper proposes a visual importance-based cross-layer 3D model streaming method over lossy network. Before the transmission, we build an important criterion for each packet of a 3D model from the aspect of rendering dependence among packets. Thus, each packet will have a rendering importance value. And then, the progressive model is streamed by...

In this paper, we propose a visual importantdriven interactive rendering method for 3D model over 802.11 WLAN for overcoming the shortcomings of wireless network's narrow bandwidth, high transmission error rates and mobile devices' low power supply. This paper first proposes an efficient simplification method based on an improved visual important r...

This paper proposes an improved UDP-lite protocol for transmitting the 3D model over wireless network. Before the transmission, the model is divided into important data and less important data. Also, we will present a importance criteria for the mesh data. And then, they are transmitted by modified protocol based on the UDP-lite respectively accord...

Retrieval relevance feedback is an iterative search technique to bridge the semantic gap between the high level user intention and low level data representation. This technique interactively determines a user's desired output or query concept by asking the user whether certain proposed 3D models are relevant or not. In the past, most research effor...

In this paper, we propose a salient region detection method for textured 3D models. We define the salient region, not only based on geometry properties, such as mean curvature, but also based on texture properties, such as texture color. Feature maps of geometry and texture are combined to form a saliency map for capturing the visually important re...

As mobile game grows rapidly in recent years, it becomes one of the new economic contributors to the game market. However the low-powered mobile device is the main barrier for the quick development of mobile game. Through mobile billiards game on the e868 mobile phone of the Bird-company, a game development framework for low performance mobile devi...

In order to overcome the shortcomings brought by the narrow bandwidth of wireless network and limited computing ability of mobile terminals, this paper mainly proposes a 3D model transmission and interactive real-time rendering method over wireless local area network. Firstly, an optimized model simplified algorithm is presented for preserve the vi...

In order to resolve the adverse visual effects brought by the lost packets over lossy wireless network, this paper mainly proposes a transmission mechanism for textured progressive model based on predictive reconstruction method. Firstly, an improved hybrid transmission technique for message packets is presented to send model information rapidly an...

In the article, we propose a adaptive FEC algorithm implemented in the Access Point (AP) to improve the quality of progressively compressed 3D models transmission over wireless networks(such as IEEE 802.11).Most FEC-based error control algorithms, which add redundant data to transmission data in a fixed number. However, most are unable to successfu...

This paper proposes a model streaming mechanism for colored progressive model based on predictive reconstruction method. After the packetizating of progressive mode and sent at the server side, predictive reconstruction method is developed to help a 3D model render in a desired visual quality even if some model information is lost. In particularly,...

The diversity of the geospatial data format has led to the difficulties in data sharing. In this paper, three popular modes for integrating heterogeneous spatial data are analyzed. Aiming at their weak extensibility, a method of exchanging data format based on middleware technology is proposed. What's more, this method can be extended freely. By es...

In order to solve the problems for lacking of both realistic for the goods presentation and the friendly interaction with user on the traditional E-commerce Websites, this paper presents applying X3D techniques into E-commerce website and realizes a 3D E-commerce shopping system. Firstly, this paper presents two key technologies employed in this sy...

When a 3D model is transmitted over a lossy network, some model information may inevitably be missing. Under such situation, one may not be able to visualize the receiving model unless the lost model information has been retransmitted. Progressive model transmission offers an alternative to avoid the “all or nothing situation” by allowing a model t...

This paper proposes a novel error resilient packetization scheme for transmitting progressive meshes over lossy networks. In the scheme, we first construct a non-redundant directed acyclic graph encoding the dependencies among the whole vertex splits of the progressive mesh. Then, it is followed by running a global graph equipartition packing algor...

A novel caching system architecture for mobile streaming was proposed, which was named 2CMSA (two-level cache mobile streaming architecture). In this system, limitations for mobile streaming system such as small memory size in mobile terminal and low bandwidth in wireless access network were avoided. According to the 2CMSA, a scheduling algorithm o...