Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ Computational Visual...arrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
Computational Visual Media
Article . 2022 . Peer-reviewed
License: CC BY
Data sources: Crossref
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
Computational Visual Media
Article
License: CC BY
Data sources: UnpayWall
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
Computational Visual Media
Article . 2021
Data sources: DOAJ
https://dx.doi.org/10.60692/2f...
Other literature type . 2021
Data sources: Datacite
https://dx.doi.org/10.60692/5p...
Other literature type . 2021
Data sources: Datacite
versions View all 6 versions
addClaim

This Research product is the result of merged Research products in OpenAIRE.

You have already added 0 works in your ORCID record related to the merged Research product.

Unsupervised random forest for affinity estimation

غابة عشوائية غير خاضعة للإشراف لتقدير التقارب
Authors: Yunai Yi; Diya Sun; Peixin Li; Taekyun Kim; Tianmin Xu; Yuru Pei;

Unsupervised random forest for affinity estimation

Abstract

Résumé Cet article présente une métrique basée sur la forêt aléatoire de regroupement non supervisée pour l'estimation de l'affinité dans des données de grande taille et de grande dimension. Le critère utilisé pour la division des nœuds pendant la construction forestière peut gérer la déficience de rang lors de la mesure de la compacité des grappes. La métrique basée sur la forêt binaire est étendue à des métriques continues en exploitant à la fois le chemin de traversée commun et le plus petit nœud parent partagé. La mesure basée sur la forêt proposée estime efficacement l'affinité en transmettant des paires de données dans la forêt à l'aide d'un nombre limité d'arbres de décision. Un algorithme de pseudo-division des feuilles (PLS) est introduit pour tenir compte des relations spatiales, ce qui régularise les mesures d'affinité et surmonte les affectations de feuilles incohérentes. La métrique basée sur la forêt aléatoire avec PLS facilite l'établissement de correspondances cohérentes et ponctuelles. Le procédé proposé a été appliqué à la reconnaissance automatique de phrases à l'aide de vidéos de couleur et de profondeur et de correspondance ponctuelle. Des expériences approfondies démontrent l'efficacité de la méthode proposée dans l'estimation de l'affinité en comparaison avec l'état de l'art.

Resumen Este documento presenta una métrica basada en bosques aleatorios de agrupación no supervisada para la estimación de afinidad en datos grandes y de alta dimensión. El criterio utilizado para la división de nodos durante la construcción forestal puede manejar la deficiencia de rango al medir la compacidad del grupo. La métrica binaria basada en bosques se extiende a métricas continuas explotando tanto la ruta transversal común como el nodo padre compartido más pequeño. La métrica basada en el bosque propuesta estima de manera eficiente la afinidad transmitiendo pares de datos en el bosque utilizando un número limitado de árboles de decisión. Se introduce un algoritmo de pseudo-división de hojas (PLS) para tener en cuenta las relaciones espaciales, que regulariza las medidas de afinidad y supera las asignaciones de hojas inconsistentes. La métrica basada en bosques aleatorios con PLS facilita el establecimiento de correspondencias consistentes y puntuales. El método propuesto se ha aplicado al reconocimiento automático de frases utilizando videos de color y profundidad y correspondencia puntual. Experimentos extensos demuestran la efectividad del método propuesto en la estimación de la afinidad en una comparación con el estado de la técnica.

Abstract This paper presents an unsupervised clustering random-forest-based metric for affinity estimation in large and high-dimensional data. The criterion used for node splitting during forest construction can handle rank-deficiency when measuring cluster compactness. The binary forest-based metric is extended to continuous metrics by exploiting both the common traversal path and the smallest shared parent node. The proposed forest-based metric efficiently estimates affinity by passing down data pairs in the forest using a limited number of decision trees. A pseudo-leaf-splitting (PLS) algorithm is introduced to account for spatial relationships, which regularizes affinity measures and overcomes inconsistent leaf assign-ments. The random-forest-based metric with PLS facilitates the establishment of consistent and point-wise correspondences. The proposed method has been applied to automatic phrase recognition using color and depth videos and point-wise correspondence. Extensive experiments demonstrate the effectiveness of the proposed method in affinity estimation in a comparison with the state-of-the-art.

ملخص تقدم هذه الورقة مقياسًا قائمًا على تجميع عشوائي للغابات غير خاضع للإشراف لتقدير التقارب في البيانات الكبيرة والعالية الأبعاد. يمكن للمعيار المستخدم لتقسيم العقدة أثناء بناء الغابات معالجة نقص الرتبة عند قياس ضغط الكتلة. يمتد المقياس الثنائي القائم على الغابات إلى المقاييس المستمرة من خلال استغلال كل من مسار العبور المشترك وأصغر عقدة أصلية مشتركة. ويقدر المقياس المقترح القائم على الغابات التقارب بكفاءة عن طريق تمرير أزواج البيانات في الغابة باستخدام عدد محدود من أشجار القرار. يتم تقديم خوارزمية تقسيم الأوراق الزائفة (PLS) لحساب العلاقات المكانية، والتي تنظم مقاييس التقارب وتتغلب على مهام الأوراق غير المتسقة. يسهل المقياس العشوائي القائم على الغابات مع PLS إنشاء مراسلات متسقة وحكيمة. تم تطبيق الطريقة المقترحة على التعرف التلقائي على العبارات باستخدام مقاطع الفيديو الملونة والعمق والمراسلات الحكيمة. تُظهر التجارب المكثفة فعالية الطريقة المقترحة في تقدير التقارب بالمقارنة مع أحدث التقنيات.

Related Organizations
Keywords

Artificial intelligence, Shape Matching, Metric (unit), Economics, Structural engineering, Node (physics), Compact space, unsupervised clustering forest, Pattern recognition (psychology), Non-negative Matrix Factorization, Cluster analysis, Engineering, Image Feature Retrieval and Recognition Techniques, Shape Matching and Object Recognition, FOS: Mathematics, Feature Selection, Data mining, affinity estimation, Spectral Clustering, Pure mathematics, QA75.5-76.95, Computer science, Tree traversal, Algorithm, Operations management, forest-based metric, Rank (graph theory), Combinatorics, Electronic computers. Computer science, Computer Science, Physical Sciences, pseudo-leaf-splitting (PLS), Computer Vision and Pattern Recognition, Face Recognition and Dimensionality Reduction Techniques, Feature Matching, Mathematics, Research Article, Random forest

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    15
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Top 10%
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Top 10%
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Top 10%
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
15
Top 10%
Top 10%
Top 10%
Green
gold