Frustratingly Easy NBNN Domain Adaptation

By Tatiana Tommasi, Barbara Caputo. 2013

95 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Representation & Self-Supervised 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 Representation & Self-Supervised Learning

Citations in the corpus

Deep Domain Confusion: Maximizing for Domain Invariance (2014, neutral): “GFK(PLS,PCA) [16] 46.4 ± 0.5 61.3 ± 0.4 66.3 ± 0.4 53.0 SA [13] 45.0 64.8 69.9 59.9 DA-NBNN [31] 52.8 ± 3.7 76.6 ± 1.7 76.2 ± 2.5 68.5 DLID [8] 51.9 78.2 89.9 73.3 DeCAF 6 S+T [11] 80.7 ± 2.3 94.8 ± 1.2 – – DaNN [14] 53.6 ± 0.2 71.2 ± 0.0 83.5 ± 0.0 69.4 Ours 84.1 ± 0.6 95.4 ± 0.4 96.3 ± 0.3 91.9…”
Domain Adaptation for Visual Applications: A Comprehensive Survey (2017, methodology): “Similarly, the methods described in [29, 171, 186, 228, 52, 243]. combine transfer learning or domain adaptation with target sample selection and labeling to iteratively update the target models.”
Return of Frustratingly Easy Domain Adaptation (2015, methodology): “DA-NBNN (Tommasi and Caputo 2013) presents an NBNN-based domain adaptation algorithm that iteratively learns a class metric while inducing a large margin separation among classes.”
Probabilistic harmonization and annotation of single‐cell transcriptomics data with deep generative models (2019, methodology): “…well established numerical methods, such as integrative non-negative matrix factorization (LIGER [18]), singular value decomposition (Scanorama [19]), or canonical correlation analysis (Seurat Alignment [13]); (ii) additional non-linear transformation of the resulting latent representations so…”
Scatter Component Analysis: A Unified Framework for Domain Adaptation and Domain Generalization (2015, unknown): “[15], [16], [29], [30], [31], [32] and object detection [33].”
ImageCLEF 2014: Overview and Analysis of the Results (2014, background): “Current experimental settings, widely used in the community, consider typically one source and one target [26], or at most two sources and one target [9,30].”
Learning Transferrable Representations for Unsupervised Domain Adaptation (2016, methodology): “The majority of the literature [13, 29, 9, 28, 32] in unsupervised domain adaptation formulates a learning problem where the task is to find a transformation matrix to align the labelled source data distribution to the unlabelled target data distribution.”
Beyond Sharing Weights for Deep Domain Adaptation (2016, unknown): “…in [42], we work in the unsuper-Beyond Accuracy A → W D → W W → D Average GFK [18] 0.464 0.613 0.663 0.530 SA [19] 0.450 0.648 0.699 0.599 DA-NBNN [40] 0.528 0.766 0.762 0.685 DLID [20] 0.519 0.782 0.899 0.733 DeCAF 6 +T [30] 0.807 0.948 - - DaNN [28] 0.536 0.712 0.835 0.694 DDC [7] 0.841 0.954…”
Learning the Roots of Visual Domain Shift (2016, background): “Some work has been recently done for dealing directly with image patches within the NBNN framework [16,17] with promising results.”
Open Set Domain Adaptation (2017, background): “Other works have also exploited greedy algorithms that iteratively add target samples to the training process, while the least relevant source samples are removed [6, 42].”