Measuring Distance between Unordered Sets of Different Sizes

Gardner, Alex; Kanno, Jinko; Duncan, Christian A.; Šelmić, Rastko R.

doi:10.1109/cvpr.2014.25

Cited by 38 publications

(35 citation statements)

References 22 publications

(19 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The 13,910 feature vectors were divided into 10 batches, and each batch is configured to distribute the six target gases uniformly in time order [22]. The MCHP dataset is comprised of five static gestures (hand poses) captured from 12 users using a Vicon motion-capture camera system and a glove with attached infrared markers on certain joints [23]. The five gestures captured were a fist, pointing with one finger, pointing with two fingers, stop (hand flat), and grab (fingers curled).…”

Section: Performance Evaluation Resultsmentioning

confidence: 99%

VLSI Implementation of Restricted Coulomb Energy Neural Network with Improved Learning Scheme

et al. 2019

View full text Add to dashboard Cite

This paper proposes a restricted coulomb energy neural network (RCE-NN) with an improved learning algorithm and presents the hardware architecture design and VLSI implementation results. The learning algorithm of the existing RCE-NN applies an inefficient radius adjustment, such as learning all neurons at the same radius or reducing the radius excessively in the learning process. Moreover, since the reliability of eliminating unnecessary neurons is estimated without considering the activation region of each neuron, it is inaccurate and leaves unnecessary neurons extant. To overcome this problem, the proposed learning algorithm divides each neuron region in the learning process and measures the reliability with different factors for each region. In addition, it applies a process of gradual radius reduction by a pre-defined reduction rate. In performance evaluations using two datasets, RCE-NN with the proposed learning algorithm showed high recognition accuracy with fewer neurons compared to existing RCE-NNs. The proposed RCE-NN processor was implemented with 197.8K logic gates in 0.535 mm 2 using a 55 nm CMOS process and operated at the clock frequency of 150 MHz.

show abstract

Section: Performance Evaluation Resultsmentioning

confidence: 99%

VLSI Implementation of Restricted Coulomb Energy Neural Network with Improved Learning Scheme

et al. 2019

View full text Add to dashboard Cite

show abstract

“…max {x i , y i } (with the subcase that x i = µ A (z) and y i = µ B (z) denote multiplicities of (occurrences of) z in multisets A and B; cf. [9]), and the Steinhaus distance [12,4] (i.e., any set measures, including probability measures). We mention that all these results can equally easily be proven by the arguments in [5]; however, for modular functions satisfying f (∅) > 0, these arguments fail.…”

Section: Remarksmentioning

confidence: 99%

A note on the triangle inequality for the Jaccard distance

Kosub

2019

Pattern Recognition Letters

182

View full text Add to dashboard Cite

Two simple proofs of the triangle inequality for the Jaccard distance in terms of nonnegative, monotone, submodular functions are given and discussed.The Jaccard index [8] is a classical similarity measure on sets with a lot of practical applications in information retrieval, data mining, machine learning, and many more (cf., e.g., [7]). Measuring the relative size of the overlap of two finite sets A and B, the Jaccard index J and the associated Jaccard distance J δ are formally defined as:

show abstract

“…The MoCap Hand Postures dataset (Gardner et al 2014) consists of 5 types of hand postures/gestures from 12 users recorded in a motion capture environment using 11 unlabeled markers attached to a glove. We only use a small subset of the data with 200 samples for each cluster.…”

Section: Mocap Hand Postures Datasetmentioning

confidence: 99%

“…In Section 5 we show that the attack is successful for different datasets where the true metric is not known. • We test our algorithm (Section 5) on Ward's Hierarchical clustering (Ward Jr 1963), and the K-Means clustering (Lloyd 1982) on multiple datasets, e.g., the UCI Handwritten Digits dataset (Alpaydin and Kaynak 1995), the MNIST dataset (LeCun 1998), the MoCap Hand Postures dataset (Gardner et al 2014), and the UCI Wheat Seeds dataset (Charytanowicz et al 2010). We find that our attack algorithm generates multiple spill-over adversarial samples across all datasets and algorithms.…”

Section: Introductionmentioning

confidence: 99%

Suspicion-Free Adversarial Attacks on Clustering Algorithms

Chhabra

Roy

Mohapatra

2020

AAAI

View full text Add to dashboard Cite

Clustering algorithms are used in a large number of applications and play an important role in modern machine learning– yet, adversarial attacks on clustering algorithms seem to be broadly overlooked unlike supervised learning. In this paper, we seek to bridge this gap by proposing a black-box adversarial attack for clustering models for linearly separable clusters. Our attack works by perturbing a single sample close to the decision boundary, which leads to the misclustering of multiple unperturbed samples, named spill-over adversarial samples. We theoretically show the existence of such adversarial samples for the K-Means clustering. Our attack is especially strong as (1) we ensure the perturbed sample is not an outlier, hence not detectable, and (2) the exact metric used for clustering is not known to the attacker. We theoretically justify that the attack can indeed be successful without the knowledge of the true metric. We conclude by providing empirical results on a number of datasets, and clustering algorithms. To the best of our knowledge, this is the first work that generates spill-over adversarial samples without the knowledge of the true metric ensuring that the perturbed sample is not an outlier, and theoretically proves the above.

show abstract

Measuring Distance between Unordered Sets of Different Sizes

Cited by 38 publications

References 22 publications

VLSI Implementation of Restricted Coulomb Energy Neural Network with Improved Learning Scheme

VLSI Implementation of Restricted Coulomb Energy Neural Network with Improved Learning Scheme

A note on the triangle inequality for the Jaccard distance

Suspicion-Free Adversarial Attacks on Clustering Algorithms

Contact Info

Product

Resources

About