Incremental Learning of NCM Forests for Large-Scale Image Classification

By Marko Ristin, Matthieu Guillaumin, Juergen Gall, Luc Van Gool. 2014

93 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Meta & Continual Learning

How researchers see this paper

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

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Meta & Continual Learning

Citations in the corpus

Towards Open Set Deep Networks (2015, neutral): “While more complex models, such as nearest class multiple centroids (NCMC) [13] or NCM forests [17], could provide more accurate modeling, for simplicity this paper focuses on just using a single mean.”
iCaRL: Incremental Classifier and Representation Learning (2016, positive): “Despite (or because of) its simplicity, NCM has been shown to work well and be more robust than standard parametric classifiers in an incremental learning setting [23, 24, 31].”
Incremental Learning in Online Scenario (2020, background): “Nearest class Mean (NCM) classiﬁer [22, 21] is a good choice where the test image is classiﬁed as the class with the closest class data mean.”
Incremental Learning of Object Detectors without Catastrophic Forgetting (2017, background): “Other work partially addresses these issues by learning classifiers without access to the ensemble of data [26, 33], but uses a fixed image representation.”
Towards Open World Recognition (2014, methodology): “This algorithm has near state of the art results and while recent extension with random forests[32] improved accuracy slightly, [32] does not provide code.”
A Strategy for an Uncompromising Incremental Learner (2017, methodology): “The NCM approach has also been successfully applied to random forest based models for incremental learning in [25].”
Recent Advances in Open Set Recognition: A Survey (2018, methodology): “Adopting the concept of Nearest Class Mean [92], [93], each class is then represented as a mean activation vector (MAV) with the mean of the activation vectors (only for the correctly classiﬁed training samples) in the penultimate layer of that network.”
IExpressNet: Facial Expression Recognition with Incremental Classes (2020, methodology): “The most classic method is the nearest class mean (NCM) classifier [45, 54].”
From categories to subcategories: Large-scale image classification with partial class label refinement (2015, methodology): “In Section 5, we show that the regularized NCMFs outperform the NCMFs [26].”
Cognitively-Inspired Model for Incremental Learning Using a Few Examples (2020, methodology): “Another set of early approaches use a ﬁxed data representation with a Nearest Class Mean (NCM) classiﬁer for incremental learning [31, 32, 41].”
Brain-Inspired Model for Incremental Learning Using a Few Examples (2020, methodology): “Another set of early approaches use a fixed data representation with a Nearest Class Mean (NCM) classifier for incremental learning [30, 31, 36].”
The Extreme Value Machine (2015, methodology): “Several methods have been proposed to deal with the incremental learning problem, but they are often either prohibitively expensive [1]–[4], or provide little to no characterization of the statistical distribution of the data [5]–[8].”
Incremental Learning Using Conditional Adversarial Networks (2019, methodology): “[29] proposed to combine the nearest neighbor classifier and random forest to form the Nearest Class Mean Forest (NCMF), and used it for incremental learning.”
Memory Based Online Learning of Deep Representations from Video Streams (2017, methodology): “The Nearest Class Mean (NCM) classifier proposed in [22], has been shown to work well and be more robust than standard parametric classifiers in an incremental learning setting [22] [23] [24].”