Authentication email has already been sent, please check your email box: and activate it as soon as possible.
You can login to My Profile and manage your email alerts.
If you haven’t received the email, please:
|
|
There are 322 papers published in subject: > since this site started. |
Select Subject |
Select/Unselect all | For Selected Papers |
Saved Papers
Please enter a name for this paper to be shown in your personalized Saved Papers list
|
1. Online Chinese Polyphone Disambiguation with Progressive Neural Networks | |||
ZHANG Yi-Fei | |||
Computer Science and Technology 16 March 2023 | |||
Show/Hide Abstract | Cite this paper︱Full-text: PDF (0 B) | |||
Abstract:There has been plenty of research on Chinese polyphonic disambiguation (CPD) problems. However, badcases are always found in real-life products. To fix such bad cases without affecting system performance on known cases is a rigid demand. In this paper, continual learning is introduced to CPD problems, and Progressive Neural Networks (PNN) is used to learn new knowledge from bad cases without sacrificing system performance on old datasets. The experimental results show that the proposed method can repair the badcase without forgetting the feature of the original dataset. Compared with the traditional finetune method, the accuracy of the model on the old dataset decreases by nearly 20\%. Our method can ensure that the accuracy of the original dataset just decreases by about 0.3\% after learning the new feature data, and the time consumption is acceptable. Potential improvements like weight pruning are also discussed. | |||
TO cite this article:ZHANG Yi-Fei. Online Chinese Polyphone Disambiguation with Progressive Neural Networks[OL].[16 March 2023] http://en.paper.edu.cn/en_releasepaper/content/4759451 |
2. A Semantic Segmentation Model for Top-Down View Image Based on Images from Multiple Vehicle On-board Cameras | |||
GE Mengcheng,SHI Yan | |||
Computer Science and Technology 10 March 2023 | |||
Show/Hide Abstract | Cite this paper︱Full-text: PDF (0 B) | |||
Abstract:Comprehensive environmental perception is crucial for autonomous driving. However, due to the issue of occlusion, current intelligent vehicle perception algorithms only recognize targets within the intelligent vehicle\'s perception area as much as possible, without predicting or annotating areas obscured by foreground objects. This limits the intelligent driving system\'s comprehensive perception and understanding of the driving environment.This paper proposes a semantic segmentation model that uses image data collected by cameras surrounding the intelligent vehicle as input. The model uses spatial transformation networks for perspective transformation and DeepLabv3p architecture as the backbone of the semantic segmentation network, which outputs the semantic segmentation perception results of the intelligent vehicle\'s driving environment from a bird\'s-eye view, including the obscured areas. In addition, this paper does not rely on manually labeled data but collects data sets through the Carla simulator and uses a designed ray-localization method for subsequent data annotation. By training on the collected data set, the proposed method achieved an MIoU score of 71.49%, which is better than traditional methods based on inverse perspective transformation and fully connected network models. | |||
TO cite this article:GE Mengcheng,SHI Yan. A Semantic Segmentation Model for Top-Down View Image Based on Images from Multiple Vehicle On-board Cameras[OL].[10 March 2023] http://en.paper.edu.cn/en_releasepaper/content/4759540 |
3. Multi-grained Location Matching with Universal Structural Coordinate Encoder for Referring Expression Grounding | |||
Yihong Zhao,Xiaojie Wang | |||
Computer Science and Technology 07 March 2023 | |||
Show/Hide Abstract | Cite this paper︱Full-text: PDF (0 B) | |||
Abstract:\justifying Referring expression grounding is a multimodal matching task involving language and vision, with the goal of locating the object in an image that is closest to the current referring expression(RE). The key to this task is not only to use the attribute of the subject in the text, but also to fully utilize the complex location information (absolute and relative location) in the image. Existing methods only encode location feature using information such as 5-dimensional coordinate and object area, which ignore some possible fine-grained clues, such as the overlap between two objects, which can be helpful in distinguishing. This paper proposes a general structure modeling approach based on mask information that is applicable to both absolute and relative location. By modeling at a fine-grained level, this paper achieves the use of the same structure for both types of location information, thereby improving modular training efficiency. Specifically, for any two objects in an image, the model extracts small-scale binary feature constructed by mask information, which correspond to the subject and object parts of the relationship, respectively. Then, it performs phrase-guided object attention on this feature and update the initial representation of the objects through multi-layer message passing to obtain cross-feature information. Conducting experiments on three of the most commonly used related datasets, results show that compared to previous methods, the model can improve the performance of modular-based referring expression grounding models in a generalizable manner, further achieving superior performance. | |||
TO cite this article:Yihong Zhao,Xiaojie Wang. Multi-grained Location Matching with Universal Structural Coordinate Encoder for Referring Expression Grounding[OL].[ 7 March 2023] http://en.paper.edu.cn/en_releasepaper/content/4759356 |
4. KDText: A Lightweight Scene Text Detector with Decoupling-Based Knowledge Distillation | |||
Lei Siyue,Yan DanFeng | |||
Computer Science and Technology 28 February 2023 | |||
Show/Hide Abstract | Cite this paper︱Full-text: PDF (0 B) | |||
Abstract:In recent years, scene text detection task has made great progress. However, most of works are devoted to improving the performance of detector and pay little attention to the practical applications. In this paper, we propose to exploit a mask branch to detect arbitrary-shaped text accurately, and compress the model to reduce computation and storage costs, achieving a balance between speed and accuracy. Specifically, we distill intermediate features and text proposal classification to transfer dark knowledge to student text detector. In the distillation process, we treat textual and Background features differently and decouple positive and negative text proposals. Experimental results on the ICDAR 2015 and ICDAR 2017 MLT datasets demonstrate the superiority of our lightweight scene text detector. | |||
TO cite this article:Lei Siyue,Yan DanFeng. KDText: A Lightweight Scene Text Detector with Decoupling-Based Knowledge Distillation[OL].[28 February 2023] http://en.paper.edu.cn/en_releasepaper/content/4759241 |
5. RKDE: Data Compression for Time-Series Data based on Kernel Density Estimation in Reservoir Algorithm . | |||
CAO Yanyun,XU Peng | |||
Computer Science and Technology 15 February 2023 | |||
Show/Hide Abstract | Cite this paper︱Full-text: PDF (0 B) | |||
Abstract:Edge-cloud collaborative anomaly detection has become the most important anomaly detection architecture. However, only in the most ideal state can the central cloud platform be fully trained with sufficient data. In the case of limited communication, we have to consider reducing the use of communication resources, but still maintain a high accuracy rate of anomaly detection. In this context, RKDE, a reservoir sampling algorithm based on kernel density estimation, is proposed to reduce the amount of data uploaded to the cloud by the edge end. By improving the probability of abnormal data being sampled, the compression pool of upload is constructed, the redundant process in gradient exchange is reduced, and the sampling process is timely fed back and adjusted according to the abnormal detection results in the cloud. At the same time, RKDE is compared with several sampling algorithms to demonstrate its performance advantages. | |||
TO cite this article:CAO Yanyun,XU Peng. RKDE: Data Compression for Time-Series Data based on Kernel Density Estimation in Reservoir Algorithm .[OL].[15 February 2023] http://en.paper.edu.cn/en_releasepaper/content/4759073 |
6. YoloDepth: Yolo with Monocular Depth Estimation for Object Distance Measurement | |||
Chen Fei-Yang,Jiao Ji-Chao | |||
Computer Science and Technology 13 February 2023 | |||
Show/Hide Abstract | Cite this paper︱Full-text: PDF (0 B) | |||
Abstract:Environmental perception system is an important part of autonomous driving. A high-precision, real-time perception system can help the vehicles make feasible decisions and reasonable plans for the next step while driving. We propose a multi-task environmental perception network (YoloDepth) that can simultaneously perform traffic object detection and distance measurement. It consists of an encoder for feature extraction and two decoders for specific tasks. Our model performs excellently on COCO 2017 object detection dataset and KITTI monocular depth estimation dataset, achieving state-of-the-art speed and accuracy, and can process both visual perception tasks simultaneously on the embedded device Jeston AGX Xavier (18.3 FPS) in real-time and maintain great accuracy. | |||
TO cite this article:Chen Fei-Yang,Jiao Ji-Chao. YoloDepth: Yolo with Monocular Depth Estimation for Object Distance Measurement[OL].[13 February 2023] http://en.paper.edu.cn/en_releasepaper/content/4759099 |
7. Pseudo-label-based Decoupling Domain Adaptation for Long-tail Distribution with Domain Discrepancy | |||
Liu YiChen,Wu ZhenYu | |||
Computer Science and Technology 13 February 2023 | |||
Show/Hide Abstract | Cite this paper︱Full-text: PDF (0 B) | |||
Abstract:In real-world scenarios, machine learning tasks suffer from long-tail distribution or domain discrepancy problems, and many recent works have proposed effective methods to solve the challenges respectively. However, few studies have paid attention to the two problems simultaneously, since long-tail distribution and domain discrepancy both perhaps influence the generalization of machine learning models. Thus, according to the upper bound error theory, a design principle is given to solve the long-tail distribution with domain discrepancy problem (LT-DD) , and a pseudo-label-based decoupling domain adaptation method (PLD-DA) is proposed following the design principle in this paper.PLD-DA follows a two-stage domain adaptation framework, which trains a domain-invariant feature extractor on the original long-tail dataset at the first stage while adjusts the classifier with reweighting method at the second stage. To improve the classification confidence for the classifier, the pseudo-label information of target domain is introduced and a self-learning strategy is used. Experiments are conducted to show that our method could achieve a well-transfered feature extractor and a confident unbiased classifier simultaneously on LT-DD tasks, improves the model's generalization compared to end-end rebalancing domain adaptation methods. | |||
TO cite this article:Liu YiChen,Wu ZhenYu. Pseudo-label-based Decoupling Domain Adaptation for Long-tail Distribution with Domain Discrepancy[OL].[13 February 2023] http://en.paper.edu.cn/en_releasepaper/content/4759091 |
8. Dynamic Embedding Networks with Contrastive Learning for Session-based Recommendation | |||
Xukai Bao,Yulong Wang | |||
Computer Science and Technology 06 February 2023 | |||
Show/Hide Abstract | Cite this paper︱Full-text: PDF (0 B) | |||
Abstract:Most recent studies on session-based recommendation (SBR) utilize RNNs or GNNs to capture complex transitions among items and employ an attention mechanism to generate session embedding. However, they fail to pay sufficient attention to the relevance of items within the session, as they simply use a fixed embedding, such as the last clicked item, the average item within the session, or a special fixed item as the user’s current interest to generate session embedding, which would allow unrelated items to influence the recommendation results. In this study, to overcome these problems, a novel \textbf{D}ynamic \textbf{E}mbedding \textbf{N}etworks (DEN), is proposed by which user preferences are dynamically and accurately captured by integrating an affinity matrix and self-attention mechanism. In addition, to avoid the overfitting problem caused by data sparsity, a self-supervised contrastive learning task is incorporated into the DEN model to enhance the representations of embeddings using only the dropout masks as noise. Extensive experimental results on three real public datasets demonstrate that the proposed method outperforms the state-of-the-art session-based recommendation methods. | |||
TO cite this article:Xukai Bao,Yulong Wang. Dynamic Embedding Networks with Contrastive Learning for Session-based Recommendation[OL].[ 6 February 2023] http://en.paper.edu.cn/en_releasepaper/content/4758949 |
9. An Image Steganalysis Model Based on Local Difference Analysis | |||
Han Ming,Jin ZhuJun,Yang Yu | |||
Computer Science and Technology 01 February 2023 | |||
Show/Hide Abstract | Cite this paper︱Full-text: PDF (0 B) | |||
Abstract:Current image steganalysis models usually extract global features to distinguish between natural image and steganographic image. However, the global features that represent the differences between images are closely related to image properties, such as embedding payload, which will mostly lead to severe performance deterioration if there is a mismatch between training and detecting sources. On the contrary, the local feature changes of the moving window in an image are mainly affected by data hiding, but not by the image properties. Therefore, a detection method based on local difference analysis is proposed in this paper. By analyzing the smoothness changes of local moving blocks, the impact of payload mismatch on steganalysis accuracy is reduced. In addition, this paper also introduces long short-term memory network technique into image steganalysis, and proposes a new detection model called ILDA-Net (Image Local Difference Analysis Net), which analyzes the changes in the local residual sequence to achieve steganalysis. Experiments on LSB and WOW steganography algorithms show that ILDA-Net can effectively reduce the impact of payload mismatch on network detection performance. | |||
TO cite this article:Han Ming,Jin ZhuJun,Yang Yu. An Image Steganalysis Model Based on Local Difference Analysis[OL].[ 1 February 2023] http://en.paper.edu.cn/en_releasepaper/content/4758921 |
10. Patch Refinement Network for Portrait Matting | |||
Zhen Cheng, Qi Qi | |||
Computer Science and Technology 25 December 2022 | |||
Show/Hide Abstract | Cite this paper︱Full-text: PDF (0 B) | |||
Abstract:Portrait matting is a challenging task in the field of computer vision, which has important applications in meeting background replacement, image editing and other scenarios. However, how to obtain high precision alpha matte, especially for the part of the edge transition area, has been a difficult problem in this field. Focusing on the post processing part of the encoder-decoder, this paper designs a new efficient patch refinement network(PRN). It first recieves a coarse alpha matte and repair the selected patches under the guidance of flaw map, these patches will be reassembled to origin alpha matte to obtain high precision results. At the same time, this paper introduces ConvGRU into the refinement layer, which can improve the temporal consistency when migrating the refinement network to the task of video portrait matting. Experiments show that the PRN allows the original image to be sent to the network after down-sampling, which can improve the performance of the network inference while recovering the high-resolution details of the alpha matte as much as possible. | |||
TO cite this article:Zhen Cheng, Qi Qi. Patch Refinement Network for Portrait Matting[OL].[25 December 2022] http://en.paper.edu.cn/en_releasepaper/content/4758745 |
Select/Unselect all | For Selected Papers |
Saved Papers
Please enter a name for this paper to be shown in your personalized Saved Papers list
|
|
About Sciencepaper Online | Privacy Policy | Terms & Conditions | Contact Us
© 2003-2012 Sciencepaper Online. unless otherwise stated