Results 31 to 40 of about 1,111,935 (267)
Improving BERT With Self-Supervised Attention
IEEE Access, 2021 One of the most popular paradigms of applying large pre-trained NLP models such as BERT is to fine-tune it on a smaller dataset. However, one challenge remains as the fine-tuned model often overfits on smaller datasets.
Yiren Chen +3 more
doaj +1 more source
LSTM-CRF Neural Network With Gated Self Attention for Chinese NER
IEEE Access, 2019 Named entity recognition (NER) is an essential part of natural language processing tasks. Chinese NER task is different from the many European languages due to the lack of natural delimiters. Therefore, Chinese Word Segmentation (CWS) is usually regarded
Yanliang Jin +5 more
doaj +1 more source
Formula Graph Self‐Attention Network for Representation‐Domain Independent Materials Discovery
Advanced Science, 2022 The success of machine learning (ML) in materials property prediction depends heavily on how the materials are represented for learning. Two dominant families of material descriptors exist, one that encodes crystal structure in the representation and the
Achintha Ihalage, Yang Hao
doaj +1 more source
SAE‐CenterNet: Self‐attention enhanced CenterNet for small dense object detection
Electronics Letters, 2023 The existing object detection models have not been sufficiently optimized for small dense objects. One of the most common solutions is to extract multi‐scale features via feature pyramid network (FPN).
Yaoshun Li, Lizhi Liu, Tongwei Lu
doaj +1 more source
Self-Attention for Speech Emotion Recognition [PDF]
Interspeech 2019, 2019 Speech Emotion Recognition (SER) has been shown to benefit from many of the recent advances in deep learning, including recurrent based and attention based neural network architectures as well. Nevertheless, performance still falls short of that of humans.
Lorenzo Tarantino +2 more
openaire +3 more sources
IEEE Access, 2022 In closed-domain Question Answering (QA), the goal is to retrieve answers to questions within a specific domain. The main challenge of closed-domain QA is to develop a model that only requires small datasets for training since large-scale corpora may not
Mahsa Abazari Kia +4 more
doaj +1 more source
Self-attentive Adversarial Stain Normalization
, 2021 Hematoxylin and Eosin (H&E) stained Whole Slide Images (WSIs) are utilized for biopsy visualization-based diagnostic and prognostic assessment of diseases. Variation in the H&E staining process across different lab sites can lead to significant variations in biopsy image appearance. These variations introduce an undesirable bias when the slides
Aman Shrivastava +9 more
openaire +5 more sources
The function of the self-attention network [PDF]
Cognitive Neuroscience, 2015 This commentary links Humphrey and Sui's proposed Self-attention Network (SAN) to the memory advantage associated with self-relevant information (i.e., the self-reference effect). Articulating this link elucidates the functional quality of the SAN in ensuring that information of potential importance to self is not lost.
openaire +2 more sources
Self-attentive Model for Headline Generation [PDF]
, 2019 Headline generation is a special type of text summarization task. While the amount of available training data for this task is almost unlimited, it still remains challenging, as learning to generate headlines for news articles implies that the model has strong reasoning about natural language.
Daniil Gavrilov +2 more
openaire +3 more sources