0%

目录

1. An Empirical Investigation of Pre-Trained Transformer Language Models for Open-Domain Dialogue Generation [PDF] 摘要
2. A Multi-Source Entity-Level Sentiment Corpus for the Financial Domain: The FinLin Corpus [PDF] 摘要
3. Sentence Analogies: Exploring Linguistic Relationships and Regularities in Sentence Embeddings [PDF] 摘要
4. Shallow Discourse Annotation for Chinese TED Talks [PDF] 摘要
5. Keeping it simple: Implementation and performance of the proto-principle of adaptation and learning in the language sciences [PDF] 摘要
6. Pseudo Labeling and Negative Feedback Learning for Large-scale Multi-label Domain Classification [PDF] 摘要
7. Investigating the Decoders of Maximum Likelihood Sequence Models: A Look-ahead Approach [PDF] 摘要
8. The growing echo chamber of social media: Measuring temporal and social contagion dynamics for over 150 languages on Twitter for 2009--2020 [PDF] 摘要
9. Multi-task Learning Based Neural Bridging Reference Resolution [PDF] 摘要
10. Discovering linguistic (ir)regularities in word embeddings through max-margin separating hyperplanes [PDF] 摘要
11. Generating Emotionally Aligned Responses in Dialogues using Affect Control Theory [PDF] 摘要
12. General-Purpose Communicative Function Recognition using a Hierarchical Network with Cascading Outputs and Maximum a Posteriori Path Estimation [PDF] 摘要
13. ECSP: A New Task for Emotion-Cause Span-Pair Extraction and Classification [PDF] 摘要
14. A Post-processing Method for Detecting Unknown Intent of Dialogue System via Pre-trained Deep Neural Network Classifier [PDF] 摘要
15. Synthetic Error Dataset Generation Mimicking Bengali Writing Pattern [PDF] 摘要
16. Natural Language QA Approaches using Reasoning with External Knowledge [PDF] 摘要
17. NYTWIT: A Dataset of Novel Words in the New York Times [PDF] 摘要
18. TEDL: A Text Encryption Method Based on Deep Learning [PDF] 摘要
19. Overview of the CCKS 2019 Knowledge Graph Evaluation Track: Entity, Relation, Event and QA [PDF] 摘要
20. Frozen Binomials on the Web: Word Ordering and Language Conventions in Online Text [PDF] 摘要
21. Graph Neural Networks Meet Neural-Symbolic Computing: A Survey and Perspective [PDF] 摘要
阅读全文 »

目录

1. A Survey on Deep Hashing Methods [PDF] 摘要
2. Captioning Images with Novel Objects via Online Vocabulary Expansion [PDF] 摘要
3. Heterogeneity Loss to Handle Intersubject and Intrasubject Variability in Cancer [PDF] 摘要
4. Probability Weighted Compact Feature for Domain Adaptive Retrieval [PDF] 摘要
5. A Hybrid Approach for Tracking Individual Players in Broadcast Match Videos [PDF] 摘要
6. Spherical formulation of moving object geometric constraints for monocular fisheye cameras [PDF] 摘要
7. Traffic Signs Detection and Recognition System using Deep Learning [PDF] 摘要
8. Automated detection of pitting and stress corrosion cracks in used nuclear fuel dry storage canisters using residual neural networks [PDF] 摘要
9. Diverse and Admissible Trajectory Forecasting through Multimodal Context Understanding [PDF] 摘要
10. Can We Read Speech Beyond the Lips? Rethinking RoI Selection for Deep Visual Speech Recognition [PDF] 摘要
11. A Neuro-AI Interface for Evaluating Generative Adversarial Networks [PDF] 摘要
12. Generalizable semi-supervised learning method to estimate mass from sparsely annotated images [PDF] 摘要
13. Noise Estimation Using Density Estimation for Self-Supervised Multimodal Learning [PDF] 摘要
14. When Deep Learning Meets Data Alignment: A Review on Deep Registration Networks (DRNs) [PDF] 摘要
15. D3Feat: Joint Learning of Dense Detection and Description of 3D Local Features [PDF] 摘要
16. Demographic Bias in Presentation Attack Detection of Iris Recognition Systems [PDF] 摘要
17. Bundle Adjustment on a Graph Processor [PDF] 摘要
18. Pixel-Level Self-Paced Learning for Super-Resolution [PDF] 摘要
19. Show, Edit and Tell: A Framework for Editing Image Captions [PDF] 摘要
20. CNN-based Repetitive self-revised learning for photos' aesthetics imbalanced classification [PDF] 摘要
21. GeoConv: Geodesic Guided Convolution for Facial Action Unit Recognition [PDF] 摘要
22. DeLTra: Deep Light Transport for Projector-Camera Systems [PDF] 摘要
23. Clean-Label Backdoor Attacks on Video Recognition Models [PDF] 摘要
24. DA4AD: End-to-end Deep Attention Aware Features Aided Visual Localization for Autonomous Driving [PDF] 摘要
25. Centrality Graph Convolutional Networks for Skeleton-based Action Recognition [PDF] 摘要
26. Metric-Scale Truncation-Robust Heatmaps for 3D Human Pose Estimation [PDF] 摘要
27. Segmentation of Satellite Imagery using U-Net Models for Land Cover Classification [PDF] 摘要
28. Optimizing JPEG Quantization for Classification Networks [PDF] 摘要
29. Anysize GAN: A solution to the image-warping problem [PDF] 摘要
30. Non-linear Neurons with Human-like Apical Dendrite Activations [PDF] 摘要
31. Knowledge graph based methods for record linkage [PDF] 摘要
32. Meta-SVDD: Probabilistic Meta-Learning for One-Class Classification in Cancer Histology Images [PDF] 摘要
33. StereoNeuroBayesSLAM: A Neurobiologically Inspired Stereo Visual SLAM System Based on Direct Sparse Method [PDF] 摘要
34. Toward Adaptive Guidance: Modeling the Variety of User Behaviors in Continuous-Skill-Improving Experiences of Machine Operation Tasks [PDF] 摘要
35. Neural networks approach for mammography diagnosis using wavelets features [PDF] 摘要
36. Likelihood Regret: An Out-of-Distribution Detection Score For Variational Auto-encoder [PDF] 摘要
37. Forgetting Outside the Box: Scrubbing Deep Networks of Information Accessible from Input-Output Observations [PDF] 摘要
38. From Perspective X-ray Imaging to Parallax-Robust Orthographic Stitching [PDF] 摘要
39. A deep learning-facilitated radiomics solution for the prediction of lung lesion shrinkage in non-small cell lung cancer trials [PDF] 摘要
40. IntrA: 3D Intracranial Aneurysm Dataset for Deep Learning [PDF] 摘要
41. Generating Embroidery Patterns Using Image-to-Image Translation [PDF] 摘要
42. Practical Full Resolution Learned Lossless Image Compression [PDF] 摘要
阅读全文 »

目录

1. Distributional semantic modeling: a revised technique to train term/word vector space models applying the ontology-related approach [PDF] 摘要
2. Quality of Word Embeddings on Sentiment Analysis Tasks [PDF] 摘要
3. On the Role of Conceptualization in Commonsense Knowledge Graph Construction [PDF] 摘要
4. Practical Annotation Strategies for Question Answering Datasets [PDF] 摘要
5. Is POS Tagging Necessary or Even Helpful for Neural Dependency Parsing? [PDF] 摘要
6. Morfessor EM+Prune: Improved Subword Segmentation with Expectation Maximization and Pruning [PDF] 摘要
7. Sensitive Data Detection and Classification in Spanish Clinical Text: Experiments with BERT [PDF] 摘要
8. Improving Neural Named Entity Recognition with Gazetteers [PDF] 摘要
9. Parsing Thai Social Data: A New Challenge for Thai NLP [PDF] 摘要
10. A Corpus for Detecting High-Context Medical Conditions in Intensive Care Patient Notes Focusing on Frequently Readmitted Patients [PDF] 摘要
11. A Framework for the Computational Linguistic Analysis of Dehumanization [PDF] 摘要
12. S-APIR: News-based Business Sentiment Index [PDF] 摘要
13. EmpTransfo: A Multi-head Transformer Architecture for Creating Empathetic Dialog Systems [PDF] 摘要
14. Automatic Compilation of Resources for Academic Writing and Evaluating with Informal Word Identification and Paraphrasing System [PDF] 摘要
15. Neural Cross-Lingual Transfer and Limited Annotated Data for Named Entity Recognition in Danish [PDF] 摘要
16. What the [MASK]? Making Sense of Language-Specific BERT Models [PDF] 摘要
17. Distill, Adapt, Distill: Training Small, In-Domain Models for Neural Machine Translation [PDF] 摘要
18. Transfer Learning for Information Extraction with Limited Data [PDF] 摘要
19. End-to-End Neural Diarization: Reformulating Speaker Diarization as Simple Multi-label Classification [PDF] 摘要
阅读全文 »

算法介绍

基数排序,从左或从依次取出数列中每个数的每个数位上的数,根据取出的数,依次放到对应键值的块的最后面,然后按键值大小取出所有块,作为数列的新顺序,不断重复直到取出数列中所有数的所有数位上的数,最终的数列即为排好序的数列。
基数排序按取数方向分为两种:从左取每个数列上的数,为最高位优先(Most Significant Digit first, MSD);从右取每个数列上的数,为最低位优先(Least Significant Digit first, LSD)
下列以LSD为例。

阅读全文 »

算法介绍

快速排序,又称分割交换排序,是目前公认的最佳排序算法,核心思想是在数列中找一个数作为中介点,将比中介点小的数放到数列的左边(将不比中介点小的数放到数列的右边,两者是同时成立的,只要记住一个就好),把中介点放到相应的位置上,再对左右两边的数列进行同样操作直到两边最多包含一个数字。

视频演示

视频地址:https://www.bilibili.com/video/av18980345

阅读全文 »

算法介绍

归并排序,也叫合并排序,中心思想是将已排好序的两个或两个以上(常为两个)数列,合并成一个大的排序数列。为了保证这两个数列是已经排好序的,需要将数列对半分,直到该数列中只包含一个数字。

视频演示

视频地址:https://www.bilibili.com/video/av18980253

python 代码

阅读全文 »

算法介绍

希尔排序,用来解决插入排序的不足:每次只能将数据往后挪动一个位置,不能往后挪动到一个比较远的位置;用gap表示数据挪动的距离,在插入排序中,gap一直等于1,在最坏情况下,将第一个数挪动最后一个位置需要挪动 n-1 次,为方便理解,用“1位插入排序”表示“插入排序”。
希尔排序中先将gap设置为一个比较大的值,进行一次“gap位插入排序”,然后将gap变小,进行一次“gap位插入排序”,不断迭代,直到gap不大于0。
gap变化过程如下:首先将gap设置为 n//meta (meta通常为质数2),然后下次将gap设置为 gap//meta,不断重复直到gap不大于0。gap更多变化方法

阅读全文 »

算法介绍

插入排序,依次取数组元素,与已排序的数据进行逐一比较,找到该元素的合适位置,放下该元素,直到完成排序。

视频演示

视频地址:https://www.bilibili.com/video/av18980488

python代码

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
#-*- coding: utf-8 -*-

# 升序
def insert_sort(data):
n = len(data)
# 外循环,依次拿出第 i 个数
for i in range(1, n):
cur_data = data[i]
j=i-1
# 内循环, 在已排序的数据中,找出当前数应该放的位置
while j>=0 and cur_data<data[j]:
data[j+1]=data[j] # 将大的数依次往后挪一位,为当前数腾位置
j-=1
data[j+1]=cur_data
return data

if __name__ == '__main__':
data = [7, 8, 5, 2, 4, 6, 3]
sorted_data = insert_sort(data.copy())
print("排序前:", data)
print("排序后:", sorted_data)
阅读全文 »