Object Classification from a Single Example Utilizing Class Relevance Metrics

By Michael Fink. 2004

287 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Few-shot Learning with Retrieval Augmented Language Models (2022, background): “Few-shot learning, the task of learning from very few examples, has been studied for decades (Thrun & Pratt, 1998; Fink, 2005; Vinyals et al., 2016), but has recently seen an explosion of interest in NLP with the arrival of large pre-trained models, which exhibit emergent few-shot learning…”
Predicting Deep Zero-Shot Convolutional Neural Networks Using Textual Descriptions (2015, background): “The concept of “Zero-Shot Learning” has been introduced in the literature [7, 8, 16, 21, 5, 17] with the aim to improve the scalability of traditional object recognition systems.”
Meta-Learning: A Survey (2018, neutral): “Earlier work on one-shot learning is largely based on hand-engineered features (Fei-Fei et al., 2006; Fei-Fei, 2006; Fink, 2005; Bart and Ullman, 2005).”
A Wholistic View of Continual Learning with Deep Neural Networks: Forgotten Lessons and the Bridge to Active and Open World Learning (2020, background): “When there is only one example with a label, it is commonly referred to as one-shot learning (Fink, 2005; Fei-Fei et al., 2006).”
Generalizing from a Few Examples (2019, methodology): “t automatically improve with experience [78]. For the thirst of learning from limited supervised information to get the hang of the task, a new machine learning problem called Few-Shot Learning (FSL) [30, 31] emerges. When there is only one exemplar to learn, FSL is also called one-shot learning pro”
Write a Classifier: Zero-Shot Learning Using Purely Textual Descriptions (2013, background): “Motivated by the practical need to learn visual classifiers of rare categories, researchers have explored approaches for learning from a single image (one-shot learning [18, 9, 11, 2]) or even from no images (zero-shot learning).”
Few-Shot Learning via Saliency-Guided Hallucination of Samples (2019, background): “One- and Few-shot Learning has been studied widely in computer vision in both shallow [26, 24, 9, 2, 8, 22] and deep learning scenarios [15, 36, 33, 10, 33, 34].”
Learning to Learn: Model Regression Networks for Easy Small Sample Learning (2016, background): “In particular, inter-class transfer [40] and crossgeneralization [7] are achieved by discovering shared feature representations: 1) captured by linear or nonlinear feature transformations [14, 63, 85, 31, 1, 94, 48], 2) obtained by feature selection [59, 60, 27] or regularization [37], 3) described ”
Transfer learning for image classification with sparse prototype representations (2008, methodology): “In the context of discriminative (maximum margin) object models Fink [9] developed a method that learns distance metrics from related problems.”
AdaDurIAN: Few-shot Adaptation for Neural Text-to-Speech with DurIAN (2020, background): “To scale the capacity for new speakers, we can adapt existing pre-trained multi-speaker sys-tem to generate new speakers’ voice, which is a well-studied subject of few-shot learning [13, 14] also known as speaker adaptation [15, 16].”
Tabula rasa: Model transfer for object category detection (2011, methodology): “[8] for one shot learning, metric-learning by Fink [11], and hierarchical classification by Zweig and Weinshall [30].”
Evaluating knowledge transfer and zero-shot learning in a large-scale setting (2011, background): “While most work based on similarity between classes [1, 10 ] require a few training samples for new classes, we employ our unsupervised approach [22] where class similarities are mined automatically using semantic relatedness measures with linguistic knowledge bases like Wikipedia or web search.”
Few-shot Sequence Learning with Transformers (2020, background): “The problem of learning a classifier from a handful examples has received considerable attention in the vision domain under the name of few-shot learning (Fink, 2005; Fei-Fei et al., 2006).”
Similarity Scores Based on Background Samples (2009, background): “The OSS draws its motivation from the growing number of so called “One-Shot Learning” techniques; that is, methods which learn from one or few training examples (see for example [12, 13 ]).”
What you saw is not what you get: Domain adaptation using asymmetric kernel transforms (2011, background): “[20], [14]), we transfer the structure of the domain shift, including changes in representation, rather than structures common to related categories.”