Intelligent sorting is an important prerequisite for the full quantitative consumption and harmless disposal of kitchen waste.The existing object detection method based on an ImageNet pre-trained model is an effective...Intelligent sorting is an important prerequisite for the full quantitative consumption and harmless disposal of kitchen waste.The existing object detection method based on an ImageNet pre-trained model is an effective way of sorting.Owing to significant domain gaps between natural images and kitchen waste images,it is difficult to reflect the characteristics of diverse scales and dense distribution in kitchen waste based on an ImageNet pre-trained model,leading to poor generalisation.In this article,the authors propose the first pre-trained model for kitchen waste sorting called KitWaSor,which combines both contrastive learning(CL)and masked image modelling(MIM)through self-supervised learning(SSL).First,to address the issue of diverse scales,the authors propose a mixed masking strategy by introducing an incomplete masking branch based on the original random masking branch.It prevents the complete loss of small-scale objects while avoiding excessive leakage of large-scale object pixels.Second,to address the issue of dense distribution,the authors introduce semantic consistency constraints on the basis of the mixed masking strategy.That is,object semantic reasoning is performed through semantic consistency constraints to compensate for the lack of contextual information.To train KitWaSor,the authors construct the first million-level kitchen waste dataset across seasonal and regional distributions,named KWD-Million.Extensive experiments show that KitWaSor achieves state-of-the-art(SOTA)performance on the two most relevant downstream tasks for kitchen waste sorting(i.e.image classification and object detection),demonstrating the effectiveness of the proposed KitWaSor.展开更多
Self-supervised learning aims to learn a universal feature representation without labels.To date,most existing self-supervised learning methods are designed and optimized for image classification.These pre-trained mod...Self-supervised learning aims to learn a universal feature representation without labels.To date,most existing self-supervised learning methods are designed and optimized for image classification.These pre-trained models can be sub-optimal for dense prediction tasks due to the discrepancy between image-level prediction and pixel-level prediction.To fill this gap,we aim to design an effective,dense self-supervised learning framework that directly works at the level of pixels(or local features)by taking into account the correspondence between local features.Specifically,we present dense contrastive learning(DenseCL),which implements self-supervised learning by optimizing a pairwise contrastive(dis)similarity loss at the pixel level between two views of input images.Compared to the supervised ImageNet pre-training and other self-supervised learning methods,our self-supervised DenseCL pretraining demonstrates consistently superior performance when transferring to downstream dense prediction tasks including object detection,semantic segmentation and instance segmentation.Specifically,our approach significantly outperforms the strong MoCo-v2 by 2.0%AP on PASCAL VOC object detection,1.1%AP on COCO object detection,0.9%AP on COCO instance segmentation,3.0%mIoU on PASCAL VOC semantic segmentation and 1.8%mIoU on Cityscapes semantic segmentation.The improvements are up to 3.5%AP and 8.8%mIoU over MoCo-v2,and 6.1%AP and 6.1%mIoU over supervised counterpart with frozen-backbone evaluation protocol.展开更多
Few-shot intent detection is a practical challenge task,because new intents are frequently emerging and collecting large-scale data for them could be costly.Meta-learning,a promising technique for leveraging data from...Few-shot intent detection is a practical challenge task,because new intents are frequently emerging and collecting large-scale data for them could be costly.Meta-learning,a promising technique for leveraging data from previous tasks to enable efficient learning of new tasks,has been a popular way to tackle this problem.However,the existing meta-learning models have been evidenced to be overfitting when the meta-training tasks are insufficient.To overcome this challenge,we present a novel self-supervised task augmentation with meta-learning framework,namely STAM.Firstly,we introduce the task augmentation,which explores two different strategies and combines them to extend meta-training tasks.Secondly,we devise two auxiliary losses for integrating self-supervised learning into meta-learning to learn more generalizable and transferable features.Experimental results show that STAM can achieve consistent and considerable performance improvement to existing state-of-the-art methods on four datasets.展开更多
Robustness is a long-standing challenge for automatic speech recognition(ASR)as the applied environment of any ASR system faces much noisier speech samples than clean training corpora.However,it is impractical to anno...Robustness is a long-standing challenge for automatic speech recognition(ASR)as the applied environment of any ASR system faces much noisier speech samples than clean training corpora.However,it is impractical to annotate every types of noisy environments.In this work,we propose a novel phonetic-semantic pre-training(PSP)framework that allows a model to effectively improve the performance of ASR against practical noisy environments via seamlessly integrating pre-training,self-supervised learning,and fine-tuning.In particular,there are three fundamental stages in PSP.First,pre-train the phone-to-word transducer(PWT)to map the generated phone sequence to the target text using only unpaired text data;second,continue training the PWT on more complex data generated from an empirical phone-perturbation heuristic,in additional to self-supervised signals by recovering the tainted phones;and third,fine-tune the resultant PWT with real world speech data.We perform experiments on two real-life datasets collected from industrial scenarios and synthetic noisy datasets,which show that the PSP effectively improves the traditional ASR pipeline with relative character error rate(CER)reductions of 28.63%and 26.38%,respectively,in two real-life datasets.It also demonstrates its robustness against synthetic highly noisy speech datasets.展开更多
Medical data refers to health-related information associated with regular patient care or as part of a clinical trial program.There are many categories of such data,such as clinical imaging data,bio-signal data,electr...Medical data refers to health-related information associated with regular patient care or as part of a clinical trial program.There are many categories of such data,such as clinical imaging data,bio-signal data,electronic health records(EHR),and multi-modality medical data.With the development of deep neural networks in the last decade,the emerging pre-training paradigm has become dominant in that it has significantly improved machine learning methods′performance in a data-limited scenario.In recent years,studies of pre-training in the medical domain have achieved significant progress.To summarize these technology advancements,this work provides a comprehensive survey of recent advances for pre-training on several major types of medical data.In this survey,we summarize a large number of related publications and the existing benchmarking in the medical domain.Especially,the survey briefly describes how some pre-training methods are applied to or developed for medical data.From a data-driven perspective,we examine the extensive use of pre-training in many medical scenarios.Moreover,based on the summary of recent pre-training studies,we identify several challenges in this field to provide insights for future studies.展开更多
The majority of vision-language pre-training(VLP)models rely on pre-trained object detectors,which incur high costs and restrict the recognition of object classes.Additionally,their encoder-based structures hinder the...The majority of vision-language pre-training(VLP)models rely on pre-trained object detectors,which incur high costs and restrict the recognition of object classes.Additionally,their encoder-based structures hinder their ability to perform text generation tasks effectively.To mitigate these challenges,we propose a Detector-free Vision-and-Language Pre-training(D-VLP)model designed to bolster intermodal interaction for unified understanding and generation tasks.Our D-VLP model employs a co-modality decoder equipped with a fused multi-attention self-attention module,enhancing feature fusion and information alignment between images and text.It is pre-trained using a novel Prefix Masked Language Modeling(prefixMLM)approach,leveraging the strengths of masked language modeling and unidirectional language modeling,which enables bidirectional processing and autoregressive token generation.Extensive experiments demonstrate that D-VLP surpasses state-of-the-art models in vision-language tasks,highlighting its superior performance and adaptability across various image-text tasks with minimal adjustments.展开更多
Recently, the emergence of pre-trained models(PTMs) has brought natural language processing(NLP) to a new era. In this survey, we provide a comprehensive review of PTMs for NLP. We first briefly introduce language rep...Recently, the emergence of pre-trained models(PTMs) has brought natural language processing(NLP) to a new era. In this survey, we provide a comprehensive review of PTMs for NLP. We first briefly introduce language representation learning and its research progress. Then we systematically categorize existing PTMs based on a taxonomy from four different perspectives. Next,we describe how to adapt the knowledge of PTMs to downstream tasks. Finally, we outline some potential directions of PTMs for future research. This survey is purposed to be a hands-on guide for understanding, using, and developing PTMs for various NLP tasks.展开更多
基金National Key Research and Development Program of China,Grant/Award Number:2021YFC1910402。
文摘Intelligent sorting is an important prerequisite for the full quantitative consumption and harmless disposal of kitchen waste.The existing object detection method based on an ImageNet pre-trained model is an effective way of sorting.Owing to significant domain gaps between natural images and kitchen waste images,it is difficult to reflect the characteristics of diverse scales and dense distribution in kitchen waste based on an ImageNet pre-trained model,leading to poor generalisation.In this article,the authors propose the first pre-trained model for kitchen waste sorting called KitWaSor,which combines both contrastive learning(CL)and masked image modelling(MIM)through self-supervised learning(SSL).First,to address the issue of diverse scales,the authors propose a mixed masking strategy by introducing an incomplete masking branch based on the original random masking branch.It prevents the complete loss of small-scale objects while avoiding excessive leakage of large-scale object pixels.Second,to address the issue of dense distribution,the authors introduce semantic consistency constraints on the basis of the mixed masking strategy.That is,object semantic reasoning is performed through semantic consistency constraints to compensate for the lack of contextual information.To train KitWaSor,the authors construct the first million-level kitchen waste dataset across seasonal and regional distributions,named KWD-Million.Extensive experiments show that KitWaSor achieves state-of-the-art(SOTA)performance on the two most relevant downstream tasks for kitchen waste sorting(i.e.image classification and object detection),demonstrating the effectiveness of the proposed KitWaSor.
文摘Self-supervised learning aims to learn a universal feature representation without labels.To date,most existing self-supervised learning methods are designed and optimized for image classification.These pre-trained models can be sub-optimal for dense prediction tasks due to the discrepancy between image-level prediction and pixel-level prediction.To fill this gap,we aim to design an effective,dense self-supervised learning framework that directly works at the level of pixels(or local features)by taking into account the correspondence between local features.Specifically,we present dense contrastive learning(DenseCL),which implements self-supervised learning by optimizing a pairwise contrastive(dis)similarity loss at the pixel level between two views of input images.Compared to the supervised ImageNet pre-training and other self-supervised learning methods,our self-supervised DenseCL pretraining demonstrates consistently superior performance when transferring to downstream dense prediction tasks including object detection,semantic segmentation and instance segmentation.Specifically,our approach significantly outperforms the strong MoCo-v2 by 2.0%AP on PASCAL VOC object detection,1.1%AP on COCO object detection,0.9%AP on COCO instance segmentation,3.0%mIoU on PASCAL VOC semantic segmentation and 1.8%mIoU on Cityscapes semantic segmentation.The improvements are up to 3.5%AP and 8.8%mIoU over MoCo-v2,and 6.1%AP and 6.1%mIoU over supervised counterpart with frozen-backbone evaluation protocol.
基金the National Natural Science Foundation of China under Grant Nos.61936012 and 61976114。
文摘Few-shot intent detection is a practical challenge task,because new intents are frequently emerging and collecting large-scale data for them could be costly.Meta-learning,a promising technique for leveraging data from previous tasks to enable efficient learning of new tasks,has been a popular way to tackle this problem.However,the existing meta-learning models have been evidenced to be overfitting when the meta-training tasks are insufficient.To overcome this challenge,we present a novel self-supervised task augmentation with meta-learning framework,namely STAM.Firstly,we introduce the task augmentation,which explores two different strategies and combines them to extend meta-training tasks.Secondly,we devise two auxiliary losses for integrating self-supervised learning into meta-learning to learn more generalizable and transferable features.Experimental results show that STAM can achieve consistent and considerable performance improvement to existing state-of-the-art methods on four datasets.
文摘Robustness is a long-standing challenge for automatic speech recognition(ASR)as the applied environment of any ASR system faces much noisier speech samples than clean training corpora.However,it is impractical to annotate every types of noisy environments.In this work,we propose a novel phonetic-semantic pre-training(PSP)framework that allows a model to effectively improve the performance of ASR against practical noisy environments via seamlessly integrating pre-training,self-supervised learning,and fine-tuning.In particular,there are three fundamental stages in PSP.First,pre-train the phone-to-word transducer(PWT)to map the generated phone sequence to the target text using only unpaired text data;second,continue training the PWT on more complex data generated from an empirical phone-perturbation heuristic,in additional to self-supervised signals by recovering the tainted phones;and third,fine-tune the resultant PWT with real world speech data.We perform experiments on two real-life datasets collected from industrial scenarios and synthetic noisy datasets,which show that the PSP effectively improves the traditional ASR pipeline with relative character error rate(CER)reductions of 28.63%and 26.38%,respectively,in two real-life datasets.It also demonstrates its robustness against synthetic highly noisy speech datasets.
基金supported by 2021 UQ School of Information Technology and Electrical Engineering(ITEE)Research Support Funding,Cyber Research Seed Funding(No.2021-R3)the University of Adelaide(No.1531570)New Staff Research Start-up Funds(No.NS-2102).
文摘Medical data refers to health-related information associated with regular patient care or as part of a clinical trial program.There are many categories of such data,such as clinical imaging data,bio-signal data,electronic health records(EHR),and multi-modality medical data.With the development of deep neural networks in the last decade,the emerging pre-training paradigm has become dominant in that it has significantly improved machine learning methods′performance in a data-limited scenario.In recent years,studies of pre-training in the medical domain have achieved significant progress.To summarize these technology advancements,this work provides a comprehensive survey of recent advances for pre-training on several major types of medical data.In this survey,we summarize a large number of related publications and the existing benchmarking in the medical domain.Especially,the survey briefly describes how some pre-training methods are applied to or developed for medical data.From a data-driven perspective,we examine the extensive use of pre-training in many medical scenarios.Moreover,based on the summary of recent pre-training studies,we identify several challenges in this field to provide insights for future studies.
基金supported in part by the Science and Technology Major Project of Guangxi under Grant No.AA22068057the National Natural Science Foundation of China under Grant No.62076077the School Foundation of Guilin University of Aerospace Technology under Grant No.XJ21KT32.
文摘The majority of vision-language pre-training(VLP)models rely on pre-trained object detectors,which incur high costs and restrict the recognition of object classes.Additionally,their encoder-based structures hinder their ability to perform text generation tasks effectively.To mitigate these challenges,we propose a Detector-free Vision-and-Language Pre-training(D-VLP)model designed to bolster intermodal interaction for unified understanding and generation tasks.Our D-VLP model employs a co-modality decoder equipped with a fused multi-attention self-attention module,enhancing feature fusion and information alignment between images and text.It is pre-trained using a novel Prefix Masked Language Modeling(prefixMLM)approach,leveraging the strengths of masked language modeling and unidirectional language modeling,which enables bidirectional processing and autoregressive token generation.Extensive experiments demonstrate that D-VLP surpasses state-of-the-art models in vision-language tasks,highlighting its superior performance and adaptability across various image-text tasks with minimal adjustments.
基金the National Natural Science Foundation of China(Grant Nos.61751201 and 61672162)the Shanghai Municipal Science and Technology Major Project(Grant No.2018SHZDZX01)and ZJLab。
文摘Recently, the emergence of pre-trained models(PTMs) has brought natural language processing(NLP) to a new era. In this survey, we provide a comprehensive review of PTMs for NLP. We first briefly introduce language representation learning and its research progress. Then we systematically categorize existing PTMs based on a taxonomy from four different perspectives. Next,we describe how to adapt the knowledge of PTMs to downstream tasks. Finally, we outline some potential directions of PTMs for future research. This survey is purposed to be a hands-on guide for understanding, using, and developing PTMs for various NLP tasks.
