FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3005087123> ?p ?o ?g. }

• W3005087123 abstract "Large facial variations are the main challenge in face recognition. To this end, previous variation-specific methods make full use of task-related prior to design special network losses, which are typically not general among different tasks and scenarios. In contrast, the existing generic methods focus on improving the feature discriminability to minimize the intra-class distance while maximizing the interclass distance, which perform well on easy samples but fail on hard samples. To improve the performance on those hard samples for general tasks, we propose a novel Distribution Distillation Loss to narrow the performance gap between easy and hard samples, which is a simple, effective and generic for various types of facial variations. Specifically, we first adopt state-of-the-art classifiers such as ArcFace to construct two similarity distributions: teacher distribution from easy samples and student distribution from hard samples. Then, we propose a novel distribution-driven loss to constrain the student distribution to approximate the teacher distribution, which thus leads to smaller overlap between the positive and negative pairs in the student distribution. We have conducted extensive experiments on both generic large-scale face benchmarks and benchmarks with diverse variations on race, resolution and pose. The quantitative results demonstrate the superiority of our method over strong baselines, e.g., Arcface and Cosface." @default.
• W3005087123 created "2020-02-14" @default.
• W3005087123 creator A5000487606 @default.
• W3005087123 creator A5006016056 @default.
• W3005087123 creator A5016080094 @default.
• W3005087123 creator A5020842034 @default.
• W3005087123 creator A5029021362 @default.
• W3005087123 creator A5045678161 @default.
• W3005087123 creator A5056674934 @default.
• W3005087123 creator A5067413001 @default.
• W3005087123 date "2020-02-10" @default.
• W3005087123 modified "2023-09-24" @default.
• W3005087123 title "Distribution Distillation Loss: Generic Approach for Improving Face Recognition from Hard Samples" @default.
• W3005087123 cites W1509966554 @default.
• W3005087123 cites W1821462560 @default.
• W3005087123 cites W1964246740 @default.
• W3005087123 cites W1998808035 @default.
• W3005087123 cites W2054515210 @default.
• W3005087123 cites W2109002610 @default.
• W3005087123 cites W2114380981 @default.
• W3005087123 cites W2145287260 @default.
• W3005087123 cites W2187089797 @default.
• W3005087123 cites W2325939864 @default.
• W3005087123 cites W2330462487 @default.
• W3005087123 cites W2510798689 @default.
• W3005087123 cites W2520774990 @default.
• W3005087123 cites W2547446130 @default.
• W3005087123 cites W2620998106 @default.
• W3005087123 cites W2731516819 @default.
• W3005087123 cites W2736633948 @default.
• W3005087123 cites W2737047298 @default.
• W3005087123 cites W2781292787 @default.
• W3005087123 cites W2792481260 @default.
• W3005087123 cites W2808174311 @default.
• W3005087123 cites W2871667416 @default.
• W3005087123 cites W2884928667 @default.
• W3005087123 cites W2903396356 @default.
• W3005087123 cites W2962982906 @default.
• W3005087123 cites W2963351448 @default.
• W3005087123 cites W2963460857 @default.
• W3005087123 cites W2963516811 @default.
• W3005087123 cites W2963676087 @default.
• W3005087123 cites W2963839617 @default.
• W3005087123 cites W2963896050 @default.
• W3005087123 cites W2985817549 @default.
• W3005087123 cites W3101998545 @default.
• W3005087123 hasPublicationYear "2020" @default.
• W3005087123 type Work @default.
• W3005087123 sameAs 3005087123 @default.
• W3005087123 citedByCount "1" @default.
• W3005087123 countsByYear W30050871232020 @default.
• W3005087123 crossrefType "posted-content" @default.
• W3005087123 hasAuthorship W3005087123A5000487606 @default.
• W3005087123 hasAuthorship W3005087123A5006016056 @default.
• W3005087123 hasAuthorship W3005087123A5016080094 @default.
• W3005087123 hasAuthorship W3005087123A5020842034 @default.
• W3005087123 hasAuthorship W3005087123A5029021362 @default.
• W3005087123 hasAuthorship W3005087123A5045678161 @default.
• W3005087123 hasAuthorship W3005087123A5056674934 @default.
• W3005087123 hasAuthorship W3005087123A5067413001 @default.
• W3005087123 hasConcept C110121322 @default.
• W3005087123 hasConcept C119857082 @default.
• W3005087123 hasConcept C120665830 @default.
• W3005087123 hasConcept C121332964 @default.
• W3005087123 hasConcept C127413603 @default.
• W3005087123 hasConcept C134306372 @default.
• W3005087123 hasConcept C138885662 @default.
• W3005087123 hasConcept C144024400 @default.
• W3005087123 hasConcept C153180895 @default.
• W3005087123 hasConcept C154945302 @default.
• W3005087123 hasConcept C178790620 @default.
• W3005087123 hasConcept C185592680 @default.
• W3005087123 hasConcept C192209626 @default.
• W3005087123 hasConcept C199360897 @default.
• W3005087123 hasConcept C201995342 @default.
• W3005087123 hasConcept C204030448 @default.
• W3005087123 hasConcept C2776401178 @default.
• W3005087123 hasConcept C2779304628 @default.
• W3005087123 hasConcept C2780451532 @default.
• W3005087123 hasConcept C2780801425 @default.
• W3005087123 hasConcept C33923547 @default.
• W3005087123 hasConcept C36289849 @default.
• W3005087123 hasConcept C41008148 @default.
• W3005087123 hasConcept C41895202 @default.
• W3005087123 hasConceptScore W3005087123C110121322 @default.
• W3005087123 hasConceptScore W3005087123C119857082 @default.
• W3005087123 hasConceptScore W3005087123C120665830 @default.
• W3005087123 hasConceptScore W3005087123C121332964 @default.
• W3005087123 hasConceptScore W3005087123C127413603 @default.
• W3005087123 hasConceptScore W3005087123C134306372 @default.
• W3005087123 hasConceptScore W3005087123C138885662 @default.
• W3005087123 hasConceptScore W3005087123C144024400 @default.
• W3005087123 hasConceptScore W3005087123C153180895 @default.
• W3005087123 hasConceptScore W3005087123C154945302 @default.
• W3005087123 hasConceptScore W3005087123C178790620 @default.
• W3005087123 hasConceptScore W3005087123C185592680 @default.
• W3005087123 hasConceptScore W3005087123C192209626 @default.
• W3005087123 hasConceptScore W3005087123C199360897 @default.
• W3005087123 hasConceptScore W3005087123C201995342 @default.
• W3005087123 hasConceptScore W3005087123C204030448 @default.

• next