Bilingual lexicon induction focuses on learning word translation pairs,also known as bitexts,from monolingual corpora by establishing a mapping between the source and target embedding spaces.Despite recent advancement...Bilingual lexicon induction focuses on learning word translation pairs,also known as bitexts,from monolingual corpora by establishing a mapping between the source and target embedding spaces.Despite recent advancements,bilingual lexicon induction is limited to inducing bitexts consisting of individual words,lacking the ability to handle semantics-rich phrases.To bridge this gap and support downstream cross-lingual tasks,it is practical to develop a method for bilingual phrase induction that extracts bilingual phrase pairs from monolingual corpora without relying on cross-lingual knowledge.In this paper,the authors propose a novel phrase embedding training method based on the skip-gram structure.Specifically,a local hard negative sampling strategy that utilises negative samples of central tokens in sliding windows to enhance phrase embedding learning is introduced.The proposed method achieves competitive or superior performance compared to baseline approaches,with exceptional results recorded for distant languages.Additionally,we develop a phrase representation learning method that leverages multilingual pre-trained language models.These mPLMs-based representations can be combined with the above-mentioned static phrase embeddings to further improve the accuracy of the bilingual phrase induction task.We manually construct a dataset of bilingual phrase pairs and integrate it with MUSE to facilitate the bilingual phrase induction task.展开更多
Let B^pΩ, 1 ≤ p 〈 ∞, be the space of all bounded functions from Lp（R） which can be extended to entire functions of exponential type Ω. The uniform error bounds for truncated Whittaker-Kotelnikov-Shannon series ...Let B^pΩ, 1 ≤ p 〈 ∞, be the space of all bounded functions from Lp（R） which can be extended to entire functions of exponential type Ω. The uniform error bounds for truncated Whittaker-Kotelnikov-Shannon series based on local sampling are derived for functions f ∈ B^pΩ without decay assumption at infinity. Then the optimal bounds of the aliasing error and truncation error of Whittaker-Kotelnikov-Shannon expansion for non-bandlimited functions from Sobolev classes L/（Wp（R）） are determined up to a logarithmic factor.展开更多
A new matting algorithm based on color distance and differential distance is proposed to deal with the problem that many matting methods perform poorly with complex natural images.The proposed method combines local sa...A new matting algorithm based on color distance and differential distance is proposed to deal with the problem that many matting methods perform poorly with complex natural images.The proposed method combines local sampling with global sampling to select foreground and background pairs for unknown pixels and then a new cost function is constructed based on color distance and differential distance to further optimize the selected sample pairs.Finally,a quadratic objective function is used based on matte Laplacian coming from KNN matting which is added with texture feature.Through experiments on various test images,it is confirmed that the results obtained by the proposed method are more accurate than those obtained by traditional methods.The four-error-metrics comparison on benchmark dataset among several algorithms also proves the effectiveness of the proposed method.展开更多
Background： Currently, the common and feasible way to estimate the most accurate forest biomass requires ground measurements and allometric models.Previous studies have been conducted on allometric equations developm...Background： Currently, the common and feasible way to estimate the most accurate forest biomass requires ground measurements and allometric models.Previous studies have been conducted on allometric equations development for estimating tree aboveground biomass（AGB） of tropical dipterocarp forests（TDFs） in Kalimantan（Indonesian Borneo）.However, before the use of existing equations, a validation for the selection of the best allometric equation is required to assess the model bias and precision.This study aims at evaluating the validity of local and pantropical equations; developing new allometric equations for estimating tree AGB in TDFs of Kalimantan; and validating the new equations using independent datasets.Methods： We used 108 tree samples from destructive sampling to develop the allometric equations, with maximum tree diameter of 175 cm and another 109 samples from previous studies for validating our equations.We performed ordinary least squares linear regression to explore the relationship between the AGB and the predictor variables in the natural logarithmic form.Results： This study found that most of the existing local equations tended to be biased and imprecise, with mean relative error and mean absolute relative error more than 0.1 and 0.3, respectively.We developed new allometric equations for tree AGB estimation in the TDFs of Kalimantan.Through a validation using an independent dataset,we found that our equations were reliable in estimating tree AGB in TDF.The pantropical equation, which includes tree diameter, wood density and total height as predictor variables performed only slightly worse than our new models.Conclusions： Our equations improve the precision and reduce the bias of AGB estimates of TDFs.Local models developed from small samples tend to systematically bias.A validation of existing AGB models is essential before the use of the models.展开更多
基金National Key Research and Development Program of China,Grant/Award Number:2023YFC3305003National Natural Science Foundation of China,Grant/Award Number:62376076。
文摘Bilingual lexicon induction focuses on learning word translation pairs,also known as bitexts,from monolingual corpora by establishing a mapping between the source and target embedding spaces.Despite recent advancements,bilingual lexicon induction is limited to inducing bitexts consisting of individual words,lacking the ability to handle semantics-rich phrases.To bridge this gap and support downstream cross-lingual tasks,it is practical to develop a method for bilingual phrase induction that extracts bilingual phrase pairs from monolingual corpora without relying on cross-lingual knowledge.In this paper,the authors propose a novel phrase embedding training method based on the skip-gram structure.Specifically,a local hard negative sampling strategy that utilises negative samples of central tokens in sliding windows to enhance phrase embedding learning is introduced.The proposed method achieves competitive or superior performance compared to baseline approaches,with exceptional results recorded for distant languages.Additionally,we develop a phrase representation learning method that leverages multilingual pre-trained language models.These mPLMs-based representations can be combined with the above-mentioned static phrase embeddings to further improve the accuracy of the bilingual phrase induction task.We manually construct a dataset of bilingual phrase pairs and integrate it with MUSE to facilitate the bilingual phrase induction task.
基金Supported by the National Natural Science Foundation of China (10971251, 11101220 and 11271199)the Program for new century excellent talents in University of China (NCET-10-0513)
文摘Let B^pΩ, 1 ≤ p 〈 ∞, be the space of all bounded functions from Lp（R） which can be extended to entire functions of exponential type Ω. The uniform error bounds for truncated Whittaker-Kotelnikov-Shannon series based on local sampling are derived for functions f ∈ B^pΩ without decay assumption at infinity. Then the optimal bounds of the aliasing error and truncation error of Whittaker-Kotelnikov-Shannon expansion for non-bandlimited functions from Sobolev classes L/（Wp（R）） are determined up to a logarithmic factor.
基金Supported by the National Natural Science Foundation of China(No.61133009,U1304616)
文摘A new matting algorithm based on color distance and differential distance is proposed to deal with the problem that many matting methods perform poorly with complex natural images.The proposed method combines local sampling with global sampling to select foreground and background pairs for unknown pixels and then a new cost function is constructed based on color distance and differential distance to further optimize the selected sample pairs.Finally,a quadratic objective function is used based on matte Laplacian coming from KNN matting which is added with texture feature.Through experiments on various test images,it is confirmed that the results obtained by the proposed method are more accurate than those obtained by traditional methods.The four-error-metrics comparison on benchmark dataset among several algorithms also proves the effectiveness of the proposed method.
基金the GIZ-Forclime project, a bilateral project between Indonesia and German governments, for funding the field measurements
文摘Background： Currently, the common and feasible way to estimate the most accurate forest biomass requires ground measurements and allometric models.Previous studies have been conducted on allometric equations development for estimating tree aboveground biomass（AGB） of tropical dipterocarp forests（TDFs） in Kalimantan（Indonesian Borneo）.However, before the use of existing equations, a validation for the selection of the best allometric equation is required to assess the model bias and precision.This study aims at evaluating the validity of local and pantropical equations; developing new allometric equations for estimating tree AGB in TDFs of Kalimantan; and validating the new equations using independent datasets.Methods： We used 108 tree samples from destructive sampling to develop the allometric equations, with maximum tree diameter of 175 cm and another 109 samples from previous studies for validating our equations.We performed ordinary least squares linear regression to explore the relationship between the AGB and the predictor variables in the natural logarithmic form.Results： This study found that most of the existing local equations tended to be biased and imprecise, with mean relative error and mean absolute relative error more than 0.1 and 0.3, respectively.We developed new allometric equations for tree AGB estimation in the TDFs of Kalimantan.Through a validation using an independent dataset,we found that our equations were reliable in estimating tree AGB in TDF.The pantropical equation, which includes tree diameter, wood density and total height as predictor variables performed only slightly worse than our new models.Conclusions： Our equations improve the precision and reduce the bias of AGB estimates of TDFs.Local models developed from small samples tend to systematically bias.A validation of existing AGB models is essential before the use of the models.
