Results 21 to 30 of about 514,020 (257)

Few-Shot Few-Shot Learning and the role of Spatial Attention [PDF]

2020 25th International Conference on Pattern Recognition (ICPR), 2021
Few-shot learning is often motivated by the ability of humans to learn new tasks from few examples. However, standard few-shot classification benchmarks assume that the representation is learned on a limited amount of base class data, ignoring the amount of prior knowledge that a human may have accumulated before learning new tasks.
Lifchitz, Yann, Avrithis, Yannis, Picard, Sylvaine   +2 more
openaire   +3 more sources

Prototype Completion for Few-Shot Learning

IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023
Few-shot learning aims to recognize novel classes with few examples. Pre-training based methods effectively tackle the problem by pre-training a feature extractor and then fine-tuning it through the nearest centroid based meta-learning. However, results show that the fine-tuning step makes marginal improvements.
Baoquan Zhang, Xutao Li, Yunming Ye, Shanshan Feng   +3 more
openaire   +3 more sources

Unsupervised Few-Shot Feature Learning via Self-Supervised Training

Frontiers in Computational Neuroscience, 2020
Learning from limited exemplars (few-shot learning) is a fundamental, unsolved problem that has been laboriously explored in the machine learning community.
Zilong Ji, Xiaolong Zou, Tiejun Huang, Si Wu, Si Wu   +4 more
doaj   +1 more source

Rich Semantics Improve Few-Shot Learning

Proceedings of the British Machine Vision Conference 2021, 2021
Human learning benefits from multi-modal inputs that often appear as rich semantics (e.g., description of an object's attributes while learning about it). This enables us to learn generalizable concepts from very limited visual examples. However, current few-shot learning (FSL) methods use numerical class labels to denote object classes which do not ...
Afham, Mohamed, Khan, Salman, Khan, Muhammad Haris, Naseer, Muzammal, Khan, Fahad Shahbaz   +4 more
openaire   +2 more sources

Few-Shot Partial-Label Learning [PDF]

Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence, 2021
Partial-label learning (PLL) generally focuses on inducing a noise-tolerant multi-class classifier by training on overly-annotated samples, each of which is annotated with a set of labels, but only one is the valid label. A basic promise of existing PLL solutions is that there are sufficient partial-label (PL) samples for training.
Zhao, Yunfeng, Yu, Guoxian, Liu, Lei, Yan, Zhongmin, Cui, Lizhen, Domeniconi, Carlotta   +5 more
openaire   +2 more sources

Applying a Probabilistic Network Method to Solve Business-Related Few-Shot Classification Problems

Complexity, 2021
It can be challenging to learn algorithms due to the research of business-related few-shot classification problems. Therefore, in this paper, we evaluate the classification of few-shot learning in the commercial field.
Lang Wu, Menggang Li
doaj   +1 more source

Feature Transformation Network for Few-Shot Learning

IEEE Access, 2021
Few-shot learning researches to learn a novel concept from a handful of labeled samples. Due to the small amount of training data, deep network has the risk of over-fitting. Although many previous approaches based on metric criterion can make significant
Xiaoyan Wang, Hongmei Wang, Daming Zhou
doaj   +1 more source

Few-Shot and Zero-Shot Learning for Historical Text Normalization [PDF]

, 2019
Historical text normalization often relies on small training datasets. Recent work has shown that multi-task learning can lead to significant improvements by exploiting synergies with related datasets, but there has been no systematic study of different ...
Bollmann, Marcel, Korchagina, Natalia, Søgaard, Anders   +2 more
core   +3 more sources

Heterogeneous Ensemble-Based Spike-Driven Few-Shot Online Learning

Frontiers in Neuroscience, 2022
Spiking neural networks (SNNs) are regarded as a promising candidate to deal with the major challenges of current machine learning techniques, including the high energy consumption induced by deep neural networks.
Shuangming Yang, Bernabe Linares-Barranco, Badong Chen   +2 more
doaj   +1 more source

Few-shot Learning: Methods and Applications [PDF]

ITM Web of Conferences
The Few-shot learning (FSL) approach distills meaningful features from a constrained sample set, allowing models to swiftly adjust to novel tasks and decreasing the dependency on extensive datasets. This approach leverages methods involving meta-learning,
Li Jiaxiang, Li Mingyang
doaj   +1 more source