Loss Of Semantic Information Research Articles

Long Text Truncation Algorithm Based on Label Embedding in Text Classification

The long text classification task has become a hot research topic in the field of text classification due to its long length and redundant information. At present, the common processing methods for long text data, such as the truncation method and pooling method, are prone to the problem of too many sentences or loss of contextual semantic information. To deal with these issues, we present LTTA-LE (Long Text Truncation Algorithm Based on Label Embedding in Text Classification), which consists of three key steps. Firstly, we build a pretraining prefix template and a label word mapping prefix template to obtain the label word embedding, and we realize the joint training of long text and label words. Secondly, we calculate the cosine similarity between the label word embedding and the long text embedding, and we filter the redundant information of the long text to reduce the text length. Finally, a three-stage model training architecture is introduced to effectively improve the classification performance and generalization ability of the model. We conduct comparative experiments on three public long text datasets, and the results show that LTTA-LE has an average F1 improvement of 1.0518% over other algorithms, which proves that our method can achieve satisfactory performance.

Fruit detection and positioning technology for a Camellia oleifera C. Abel orchard based on improved YOLOv4-tiny model and binocular stereo vision

In the complex environment of an orchard, changes in illumination, leaf occlusion, and fruit overlap make it challenging for mobile picking robots to detect and locate oil-seed camellia fruit. To address this problem, YOLO-Oleifera was developed as a fruit detection model method based on a YOLOv4-tiny model, To obtain clustering results appropriate to the size of the Camellia oleifera fruit, the k-means++ clustering algorithm was used instead of the k-means clustering algorithm used by the YOLOv4-tiny model to determine bounding box priors. Two convolutional kernels of 1 × 1 and 3 × 3 were respectively added after the second and third CSPBlock modules of the YOLOv4-tiny model. This model allows the learning of Camellia oleifera fruit feature information and reduces overall computational complexity. Compared with the classic stereo matching method based on binocular camera images, this method innovatively used the bounding box generated by the YOLO-Oleifera model to extract the region of interest of the fruit, and then adaptively performs stereo matching according to the generation mechanism of the bounding box. This allows the determination of disparity and facilitates the subsequent use of the triangulation principle to determine the picking position of the fruit. An ablation experiment demonstrated the effective improvement of the YOLOv4-tiny model. Camellia oleifera fruit images obtained under sunlight and shading conditions were used to test the YOLO-Oleifera model, and the model robustly detected the fruit under different illumination conditions. Occluded Camellia oleifera fruit decreased precision and recall due to the loss of semantic information. Comparison of this model with deep learning models YOLOv5-s,YOLOv3-tiny, and YOLOv4-tiny, the YOLO-Oleifera model achieved the highest AP of 0.9207 with the smallest data weight of 29 MB. The YOLO-Oleifera model took an average of 31 ms to detect each fruit image, fast enough to meet the demand for real-time detection. The algorithm exhibited high positioning stability and robust function despite changes in illumination. The results of this study can provide a technical reference for the robust detection and positioning of Camellia oleifera fruit by a mobile picking robot in a complex orchard environment.

Knowledge Graph Entity Alignment Using Relation Structural Similarity

Embedding-based entity alignment, which represents knowledge graphs as low-dimensional embeddings and finds entities in different knowledge graphs that semantically represent the same real-world entity by measuring the similarities between entity embeddings, has achieved promising results. However, existing methods are still challenged by the error accumulation of embeddings along multi-step paths and the semantic information loss. This paper proposes a novel embedding-based entity alignment method that iteratively aligns both entities and relations with high similarities as training data. Newly-aligned entities and relations are used to calibrate the corresponding embeddings in the unified embedding space, which reduces the error accumulation. To reduce the negative impact of semantic information loss, the authors propose to use relation structural similarity instead of embedding similarity to align relations. Experimental results on five widely used real-world datasets show that the proposed method significantly outperforms several state-of-the-art methods for entity alignment.

Modeling N-ary relational data as gyro-polygons with learnable gyro-centroid

Binary Knowledge base (KB) embedding has been broadly studied, while n-ary relational KB embedding which aims to simultaneously map binary and beyond-binary facts into low-dimensional vector space is less explored. Existing n-ary KB embedding approaches typically treat entities in n-ary fact as equally important or decompose n-ary relational fact into subtuples. And they generally model n-ary relational KBs in the Euclidean space. However, n-ary facts are semantically and structurally intact, decomposition leads to intrinsic semantic information loss and undermines the structural integrity. Moreover, compared to the binary relational KBs, n-ary ones are characterized by more abundant and complicated hierarchy structures, which could not be well expressed in Euclidean space and have been greatly overlooked. To address these issues, we propose a gyro-polygon embedding framework to realize n-ary fact structure keeping, semantic integrity retaining, and hierarchy capturing, termed as PolygonE. Specifically, n-ary relational facts are modeled as gyro-polygons in the hyperbolic space, where we denote entities in facts as vertexes of gyro-polygons and relations as entity translocation operations. Importantly, we design a fact plausibility measuring strategy based on the vertex–gyrocentroid geodesic to optimize the relation-adjusted gyro-polygon. We further extend PolygonE by unsupervisedly learning the weight of each entity, and strengthening the connection between the primary triple and the whole fact with the gyrocentroid–gyromidpoint distance. Extensive experiments demonstrate the superiority of PolygonE.

Learning reliable modal weight with transformer for robust RGBT tracking

Many Siamese-based RGBT trackers have been prevalently designed in recent years for fast-tracking. However, the correlation operation in them is a local linear matching process, which may easily lose semantic information required inevitably by high-precision trackers. In this paper, we propose a strong cross-modal model based on transformer for robust RGBT tracking. Specifically, a simple dual-flow convolutional network is designed to extract and fuse dual-modal features, with comparably lower complexity. Besides, to enhance the feature representation and deepen semantic features, a modal weight allocation strategy and a backbone feature extracted network based on modified Resnet-50 are designed, respectively. Also, an attention-based transformer feature fusion network is adopted to improve long-distance feature association to decrease the loss of semantic information. Finally, a classification regression subnetwork is investigated to accurately predict the state of the target. Sufficient experiments have been implemented on the RGBT234, RGBT210, GTOT and LasHeR datasets, demonstrating more outstanding tracking performance against the state-of-the-art RGBT trackers.

Documenting the Structure and Adaptive Reuse of Roman Amphitheatres through the CIDOC CRMba Model

This article addresses an important aspect of the built heritage documentation, which concerns encoding information about a building in a formal way, making it available for reuse by the research community. Formal ontologies allow structuring and integrating information from heterogeneous sources without loss of semantic information. In the field of Cultural Heritage (CH), the CIDOC Conceptual Reference Model (CRM) ontology is well known and widely accepted as it provides definitions and a formal structure to describe the implicit and explicit concepts and relationships used in the CH documentation. One of its extensions, the CRMba model, has been specifically designed to document information on a built structure and its components. In this work, we have applied the CRMba model to the documentation of Roman architectures, in particular, Roman amphitheatres, demonstrating how the semantic model allows encoding information about the structure of the building and its evolution over time and space, stressing on the concepts of “empty spaces” and “functional spaces” defined by form, and focusing on the relationship between form and function. The aim of the work is to explore the potentiality of the model and to provide, through a series of examples supported by graphs, standard encoding procedures to be reused by scholars dealing with similar case studies.

Deep Cascade Network for Noise-Robust SAR Ship Detection With Label Augmentation

Deep learning has recently made an impressive advance in ship detection in Synthetic Aperture Radar (SAR) images. In spite of this advancement, conventional deep detection networks often suffer from speckle noise that inherently occurs in SAR images. However, despeckling researches have focused only on improving the visual quality of the SAR images. Despeckling without considering subsequent task may cause loss of semantic information and result in performance degradation. In this letter, we propose a deep cascade framework for noise-robust SAR ship detection that sequentially performs despeckle and detection. We effectively train our cascade network using pseudo SAR images with SAR-like structures and additional detection annotations. We also propose semantic conservative loss that allows these two tasks to cooperate with each other. Experimental results including comparisons to previous methods and extensive ablation studies show the effectiveness of our proposed method.

Improved CenterNet for Accurate and Fast Fitting Object Detection

Accurate and fast detection of typical fittings is the prerequisite of condition monitoring and fault diagnosis. At present, most successful fitting detectors are anchor‐based, which are challenging to meet the requirements of edge deployment. In this paper, we propose a novel anchor‐free method called HRM‐CenterNet. Firstly, the lightweight MobileNetV3 is introduced into CenterNet to extract multi‐scale features of different layers. In addition, the lightweight receptive field enhancement module is proposed for the deep layer features to further enhance the characterization power of global features and generate more accurate heatmaps. Finally, the high‐resolution feature fusion network with iterative aggregation is designed to reduce the loss of spatial semantic information in subsampling and further improve the accuracy of small and occlusion objects. Experiments are carried out on the TFITS and PASCAL VOC datasets. The results show that the size of the network is more than 60% lower than that of CenterNet. Compared with other detectors, our method achieves comparable accuracy with all accurate models at a much faster speed and meets the performance requirements of real‐time detection.

Pyramid Attention Upsampling Module for Object Detection

The core task of object detection is to extract features of various sizes by hierarchically stacking multi-scale feature maps. However, it is not easy to decide whether we should transmit semantic information to the low layers while reducing the loss of semantic information of the high-level features. In this paper, we present a novel method to reduce the loss of semantic information, and at the same time to improve the object detection performance by using the attention mechanism on the high-level layer of the feature pyramid network. The proposed method focuses on the sparse spatial information using deformable convolution v2 (DCNv2) on the lateral connection in the feature pyramid network. Specifically, the upsampling process is divided into two branches. The first one pays attention to the global context information of high-level features, and the other rescales the feature map by interpolation. Finally, by multiplying the results from the two branches, we can obtain upsampling result that pays attention to semantic information of the high-level layer. The proposed pyramid attention upsampling module has three contributions. First, It can be easily applied to any models using feature pyramid network. Second, it is possible to reduce losses in semantic information of the high-level feature map by performing context attention of the high-level layer. Third, it improves the detection performance by stacking layers up to the low layer. We used MS-COCO 2017 detection dataset to evaluate the performance of the proposed method. Experimental results show that the proposed method provided better detection performance comparing with existing feature pyramid network-base methods.

3D Reconstruction of Laparoscope Images With Contrastive Learning Methods

The use of laparoscopic images and videos to reconstruct abdominal tissue structure overcomes the visual limitations of human eyes and provides great convenience for the detection and diagnosis of medical diseases. The method described in this paper is based on contrast learning and ORB-SLAM design. The contributions of this study are as follows. (i) A data preprocessing thread is introduced, which includes data augmentation and input frame evaluation. (ii) Encoding global information to avoid semantic information loss and eliminate mismatch by designing an improved U-NET network. (iii) The improved U-NET network was used to introduce a dual-branched Siamese network and GPU-accelerated intensive reconstruction thread. The dual-branched Siamese network structure was used to compute the depth information of the feature points and optical flow information in parallel, and the dense depth estimation was obtained without interrupting the original sparse reconstruction. (iv) An experimental system was established to conduct three-dimensional reconstruction tests on laparoscopic/endoscope /UBE videos of 186 patients. To effectively provide more accurate feature detection and matching support for the application of combat scenes, the influence of lens distortion was considered. Compared with the current mainstream three-dimension reconstruction and deep learning algorithms, the practicability and superiority of laparoscopic three-dimension reconstruction based on contrastive learning are demonstrated in clinical scenarios such as surgical navigation, auxiliary diagnosis, and surgical simulation.

Human Pose Estimation Based on Multi-resolution Feature Parallel Network for Public Security

Smoking detection is an essential part of safety production management. With the wide application of artificial intelligence technology in all kinds of behavior monitoring applications, the technology of real-time monitoring smoking behavior in production areas based on video is essential. In order to carry out smoking detection, it is necessary to analyze the position of key points and posture of the human body in the input image. Due to the diversity of human pose and the complex background in general scene, the accuracy of human pose estimation is not high. To predict accurate human posture information in complex backgrounds, a deep learning network is needed to obtain the feature information of different scales in the input image. The human pose estimation method based on multi-resolution feature parallel network has two parts. The first is to reduce the loss of semantic information by hole convolution and deconvolution in the part of multi-scale feature fusion. The second is to connect different resolution feature maps in the output part to generate the high-quality heat map. To solve the problem of feature loss of previous serial models, more accurate human pose estimation data can be obtained. Experiments show that the accuracy of the proposed method on the coco test set is significantly higher than that of other advanced methods. Accurate human posture estimation results can be better applied to the field of smoking detection, and the smoking behavior can be detected by artificial intelligence, and the alarm will be automatically triggered when the smoking behavior is found.

RAFNet: RGB-D attention feature fusion network for indoor semantic segmentation

Semantic segmentation based on the complementary information from RGB and depth images has recently gained great popularity, but due to the difference between RGB and depth maps, how to effectively use RGB-D information is still a problem. In this paper, we propose a novel RGB-D semantic segmentation network named RAFNet, which can selectively gather features from the RGB and depth information. Specifically, we construct an architecture with three parallel branches and propose several complementary attention modules. This structure enables a fusion branch and we add the Bi-directional Multi-step Propagation (BMP) strategy to it, which can not only retain the feature streams of the original RGB and depth branches but also fully utilize the feature flow of the fusion branch. There are three kinds of complementary attention modules that we have constructed. The RGB-D fusion module can effectively extract important features from the RGB and depth branch streams. The refinement module can reduce the loss of semantic information and the context aggregation module can help propagate and integrate information better. We train and evaluate our model on NYUDv2 and SUN-RGBD datasets, and prove that our model achieves state-of-the-art performances.

Ethical implications of digital infrastructures for pluralistic perspectives

It is important to design digital infrastructure that can better accommodate multicultural and pluralistic views from its foundations. It is insufficient to look at only the responses and influences of culture on technology without considering how the technology can be adapted in anticipation of, and to support, pluralistic multicultural perspectives in its original design. This goes beyond the simple act of supporting multiple languages and interfaces, but should include the ability of digital and data infrastructure to capture and accommodate pluralistic views, supporting multiple perspectives in the representation and processing of the data itself. In this work, we look at how rethinking the representation of data can allow us to more directly tackle domains that are typically hampered due to intercultural differences, and their inevitable losses in translation, particularly losses of valuable information like context and intention. When we refer to a loss of context and intention, we are referring to the loss of semantic information when practices such as referencing and citation are hard-coded to a particular set of cultural norms. We show that it is possible to expand the way in which we track referential data to capture richer semantic information regarding the contexts and intentions of the creators of this data, and thereby better serve the varied needs of those who consume, study, and refer to such data. We demonstrate this concept through a prototype system for a multicultural digital infrastructure, which we have named MultiVerse, and discuss its ethical implications from the perspectives of ‘multistability’, Intercultural Information Ethics framework, and poststructuralism.

Distributed Learning and Inference With Compressed Images.

Modern computer vision requires processing large amounts of data, both while training the model and/or during inference, once the model is deployed. Scenarios where images are captured and processed in physically separated locations are increasingly common (e.g. autonomous vehicles, cloud computing, smartphones). In addition, many devices suffer from limited resources to store or transmit data (e.g. storage space, channel capacity). In these scenarios, lossy image compression plays a crucial role to effectively increase the number of images collected under such constraints. However, lossy compression entails some undesired degradation of the data that may harm the performance of the downstream analysis task at hand, since important semantic information may be lost in the process. Moreover, we may only have compressed images at training time but are able to use original images at inference time (i.e. test), or vice versa, and in such a case, the downstream model suffers from covariate shift. In this paper, we analyze this phenomenon, with a special focus on vision-based perception for autonomous driving as a paradigmatic scenario. We see that loss of semantic information and covariate shift do indeed exist, resulting in a drop in performance that depends on the compression rate. In order to address the problem, we propose dataset restoration, based on image restoration with generative adversarial networks (GANs). Our method is agnostic to both the particular image compression method and the downstream task; and has the advantage of not adding additional cost to the deployed models, which is particularly important in resource-limited devices. The presented experiments focus on semantic segmentation as a challenging use case, cover a broad range of compression rates and diverse datasets, and show how our method is able to significantly alleviate the negative effects of compression on the downstream visual task.

Liver vessel segmentation based on inter-scale V-Net.

Segmentation and visualization of liver vessel is a key task in preoperative planning and computer-aided diagnosis of liver diseases. Due to the irregular structure of liver vessel, accurate liver vessel segmentation is difficult. This paper proposes a method of liver vessel segmentation based on an improved V-Net network. Firstly, a dilated convolution is introduced into the network to make the network can still enlarge the receptive field without reducing down-sampling and save detailed spatial information. Secondly, a 3D deep supervision mechanism is introduced into the network to speed up the convergence of the network and help the network learn semantic features better. Finally, inter-scale dense connections are designed in the decoder of the network to prevent the loss of high-level semantic information during the decoding process and effectively integrate multi-scale feature information. The public datasets 3Dircadb were used to perform liver vessel segmentation experiments. The average dice and sensitivity of the proposed method reached 71.6 and 75.4%, respectively, which are higher than those of the original network. The experimental results show that the improved V-Net network can automatically and accurately segment labeled or even other unlabeled liver vessels from the CT images.

Differences on autobiographical memory in healthy aging and Alzheimer’s disease

AbstractBackgroundAutobiographical memory (AM) refers to memories from one’s past that are characterized by a sense of subjective time and autonoetic consciousness which involves a feeling of re‐experiencing or reliving the past. Cognitive aging research indicates age‐related deficits in episodic AM, whereas semantic AM is unimpaired in healthy older adults. Regarding research involving subjects with dementia had showed that episodic memories deteriorate rapidly in Alzheimer's disease (AD), whereas semantic memories are preserved until moderate stages. This study aims to compare healthy older adults and AD patients, on the type of AM in order to identify differences between the groups.MethodThirty patients with initial AD, and 46 healthy older participants matched on age and education took part in the present study. Participants underwent a neurological and neuropsychological assessment. To evaluated AM, Autobiographical Memory Interview (AMI) was applied. AMI is a semi‐structured interview, applicable to subjects from 18 years old, that evaluates the two components of AM: semantic and episodic. For the study, MANOVA was performed to determine whether there were differences between groups on AM components.ResultThe multivariate contrasts revealed a main effect of group (Λ = .701; F(2, 73) = 85.57; p < .001). Therefore, post hoc Tukey tests were performed for each dependent variable, and significant differences were obtained on all the measures studied. Healthy older people obtained higher score than AD on the episodic AM (F(1, 75) = 8.14; p < .001) and on semantic AM (F(1, 75) = 164.5; p < .001).ConclusionAutobiographical recall in AD is mainly characterized by a substantial loss of episodic and semantic information, leading to a decontextualization of autobiographical memories and a shift from mentally reliving past events to a general sense of familiarity. This decline is exacerbated by anterograde and retrograde amnesia, leaving AD patients with limited access to memories that shape self‐consciousness, self‐knowledge, and self‐images.

High-resolution dermoscopy image synthesis with conditional generative adversarial networks

Background and objectiveDue to the lack of training data, the accurate classification of skin lesions still has great challenges. Generative adversarial networks (GANs) have been used to synthesize dermoscopy images successfully. Unfortunately, previous methods usually directly feed category labels into GANs and cannot provide effective information gain for the classification model. This paper studies a specific conditional image synthesis method, which could convert semantic segmentation map to dermoscopy image. MethodsWe proposed a conditional GANs (CGAN) for high-resolution dermoscopy images synthesis. First, we established an effective label mapping with pathological significance by combining the segmentation mask and category label of skin lesions. Then, a CGAN based on the image-to-image translation framework is constructed and took the previous label mapping as input to generate dermoscopy images. Especially, the shallow and deep features are combined together in the generator to avoid the loss of semantic information, and discriminator-based feature matching loss is introduced to improve the quality of generated images. ResultsThe proposed method is evaluated in ISIC-2017 skin dataset. Compared with several representative GANs architectures including the newest semantic image synthesis method, the proposed method has better performance in both visual effect and quantitative evaluation. Moreover, by using the generated images, the average AUC values of several skin lesion classification models can be improved effectively. ConclusionsThe proposed method can generate high-realistic and high-resolution demoscopy images, leading to performance improvement of skin lesion classification models, which could also be helpful for solving data shortages and classes imbalance problems in the field of medical image analysis.

Semantics-Preserving RDB2RDF Data Transformation Using Hierarchical Direct Mapping

Direct mapping is an automatic transformation method used to generate resource description framework (RDF) data from relational data. In the field of direct mapping, semantics preservation is critical to ensure that the mapping method outputs RDF data without information loss or incorrect semantic data generation. However, existing direct-mapping methods have problems that prevent semantics preservation in specific cases. For this reason, a mapping method is developed to perform a semantics-preserving transformation of relational databases (RDB) into RDF data without semantic information loss and to reduce the volume of incorrect RDF data. This research reviews cases that do not generate semantics-preserving results, and the corresponding problems into categories are arranged. This paper defines lemmas that represent the features of RDF data transformation to resolve those problems. Based on the lemmas, this work develops a hierarchical direct-mapping method to strictly abide by the definition of semantics preservation and to prevent semantic information loss, reducing the volume of incorrect RDF data generated. Experiments demonstrate the capability of the proposed method to perform semantics-preserving RDB2RDF data transformation, generating semantically accurate results. This work impacts future studies, which should involve the development of synchronization methods to achieve RDF data consistency when original RDB data are modified.

Scalable deep asymmetric hashing via unequal-dimensional embeddings for image similarity search

In recent years, Hashing has become a popular technique used to support large-scale image retrieval, due to its significantly reduced storage, high search speed and capability of mapping high dimensional original features into compact similarity-preserving binary codes. Although effectiveness achieved, most existing hashing methods are still some limitations, including: (1) Many supervised hashing methods only transform the label information into pairwise similarities to guide the hash code learning process, which will lead to the loss of rich semantic information between image pairs. (2) Some pioneer hashing methods use a relaxation-based strategy to solve discrete problems, resulting in a large quantization error. (3) Some supervised hashing methods handle the hashing learning procedure based on an asymmetric learning manner, and although this partially solves the problems of low efficiency and accuracy of symmetric learning strategy, they are all based on the embeddings of equal dimension, which leads to the reduction in the models representation ability and an increase in potential noise. To overcome the above limitations, in this paper, we propose a novel yet simple but effective hashing method, named Scalable Deep Asymmetric Hashing (SDAH). Specifically, SDAH is an end-to-end deep hashing method based on a fast iterative optimization strategy, which utilizes two real-valued embeddings of unequal dimensions, i.e., real-valued embeddings of images and labels, to flexibly perform asymmetric similarity computation. It can circumvent the use of the large pairwise similarity matrix by introducing an intermediate label matrix term which results in a remarkable reduction in the memory space cost. By doing this, the learned hash codes are more semantically informative for image retrieval tasks. Experimental results on several benchmark datasets highlight the superiority of SDAH in comparison with many state-of-the-art hashing methods.

Deep attentive and semantic preserving video summarization

Video summarization shortens a lengthy video into a succinct version, whose challenges mainly originate from the difficulties of discovering the inherent relations between the original video and its summary, meanwhile minimizing the semantic information loss. Supervised approaches, especially those in deep learning framework, have demonstrated their effectiveness in video summarization. However, these approaches mainly focus on one of the challenges, and seldom pay close attention to both challenges simultaneously. To this end, we propose to pay close attention to this deficiency by incorporating the ideas of both the encoder-decoder attention and semantic preserving loss in a deep Seq2Seq framework for video summarization. Moreover, we also introduce Huber loss to replace the popular mean square error loss to enhance the robustness of the model to outliers. Extensive experiments on two benchmark video summarization datasets demonstrate that the proposed approach consistently outperforms the state-of-the-art ones.