FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3171314215> ?p ?o ?g. }

• W3171314215 abstract "Hand pose estimation is difficult due to different environmental conditions, object- and self-occlusion as well as diversity in hand shape and appearance. Exhaustively covering this wide range of factors in fully annotated datasets has remained impractical, posing significant challenges for generalization of supervised methods. Embracing this challenge, we propose to combine ideas from adversarial training and motion modelling to tap into unlabeled videos. To this end we propose what to the best of our knowledge is the first motion model for hands and show that an adversarial formulation leads to better generalization properties of the hand pose estimator via semi-supervised training on unlabeled video sequences. In this setting, the pose predictor must produce a valid sequence of hand poses, as determined by a discriminative adversary. This adversary reasons both on the structural as well as temporal domain, effectively exploiting the spatio-temporal structure in the task. The main advantage of our approach is that we can make use of unpaired videos and joint sequence data both of which are much easier to attain than paired training data. We perform extensive evaluation, investigating essential components needed for the proposed framework and empirically demonstrate in two challenging settings that the proposed approach leads to significant improvements in pose estimation accuracy. In the lowest label setting, we attain an improvement of $40%$ in absolute mean joint error." @default.
• W3171314215 created "2021-06-22" @default.
• W3171314215 creator A5025000908 @default.
• W3171314215 creator A5056503617 @default.
• W3171314215 creator A5066945976 @default.
• W3171314215 creator A5067824015 @default.
• W3171314215 creator A5068220806 @default.
• W3171314215 date "2021-06-10" @default.
• W3171314215 modified "2023-09-27" @default.
• W3171314215 title "Adversarial Motion Modelling helps Semi-supervised Hand Pose Estimation." @default.
• W3171314215 cites W1735317348 @default.
• W3171314215 cites W1836465849 @default.
• W3171314215 cites W2099471712 @default.
• W3171314215 cites W2101032778 @default.
• W3171314215 cites W2194775991 @default.
• W3171314215 cites W2296629652 @default.
• W3171314215 cites W2593414223 @default.
• W3171314215 cites W2605973302 @default.
• W3171314215 cites W2606517404 @default.
• W3171314215 cites W2606965392 @default.
• W3171314215 cites W2768683308 @default.
• W3171314215 cites W2784473436 @default.
• W3171314215 cites W2798581336 @default.
• W3171314215 cites W2892644985 @default.
• W3171314215 cites W2896588340 @default.
• W3171314215 cites W2897765997 @default.
• W3171314215 cites W2903831537 @default.
• W3171314215 cites W2904383899 @default.
• W3171314215 cites W2955582700 @default.
• W3171314215 cites W2962811204 @default.
• W3171314215 cites W2962916650 @default.
• W3171314215 cites W2963207848 @default.
• W3171314215 cites W2963324990 @default.
• W3171314215 cites W2963389355 @default.
• W3171314215 cites W2963548793 @default.
• W3171314215 cites W2963601560 @default.
• W3171314215 cites W2963836885 @default.
• W3171314215 cites W2963995996 @default.
• W3171314215 cites W2964093990 @default.
• W3171314215 cites W2964203186 @default.
• W3171314215 cites W2964211001 @default.
• W3171314215 cites W2968722025 @default.
• W3171314215 cites W2971856312 @default.
• W3171314215 cites W2973857456 @default.
• W3171314215 cites W2975420824 @default.
• W3171314215 cites W2979577579 @default.
• W3171314215 cites W2982601397 @default.
• W3171314215 cites W2983925976 @default.
• W3171314215 cites W2984612350 @default.
• W3171314215 cites W2993797559 @default.
• W3171314215 cites W3002101995 @default.
• W3171314215 cites W3016386342 @default.
• W3171314215 cites W3021322563 @default.
• W3171314215 cites W3034470433 @default.
• W3171314215 cites W3034479523 @default.
• W3171314215 cites W3035385851 @default.
• W3171314215 cites W3035467087 @default.
• W3171314215 cites W3035551320 @default.
• W3171314215 cites W3036644940 @default.
• W3171314215 cites W3045246204 @default.
• W3171314215 cites W3048474702 @default.
• W3171314215 cites W3054000249 @default.
• W3171314215 cites W3081254563 @default.
• W3171314215 cites W3094548594 @default.
• W3171314215 cites W3101151469 @default.
• W3171314215 cites W3104876213 @default.
• W3171314215 cites W3107620906 @default.
• W3171314215 hasPublicationYear "2021" @default.
• W3171314215 type Work @default.
• W3171314215 sameAs 3171314215 @default.
• W3171314215 citedByCount "1" @default.
• W3171314215 countsByYear W31713142152021 @default.
• W3171314215 crossrefType "posted-content" @default.
• W3171314215 hasAuthorship W3171314215A5025000908 @default.
• W3171314215 hasAuthorship W3171314215A5056503617 @default.
• W3171314215 hasAuthorship W3171314215A5066945976 @default.
• W3171314215 hasAuthorship W3171314215A5067824015 @default.
• W3171314215 hasAuthorship W3171314215A5068220806 @default.
• W3171314215 hasConcept C103278499 @default.
• W3171314215 hasConcept C104114177 @default.
• W3171314215 hasConcept C105795698 @default.
• W3171314215 hasConcept C115961682 @default.
• W3171314215 hasConcept C119857082 @default.
• W3171314215 hasConcept C134306372 @default.
• W3171314215 hasConcept C153180895 @default.
• W3171314215 hasConcept C154945302 @default.
• W3171314215 hasConcept C162324750 @default.
• W3171314215 hasConcept C177148314 @default.
• W3171314215 hasConcept C185429906 @default.
• W3171314215 hasConcept C187736073 @default.
• W3171314215 hasConcept C2778112365 @default.
• W3171314215 hasConcept C2780451532 @default.
• W3171314215 hasConcept C2781238097 @default.
• W3171314215 hasConcept C33923547 @default.
• W3171314215 hasConcept C37736160 @default.
• W3171314215 hasConcept C38652104 @default.
• W3171314215 hasConcept C41008148 @default.
• W3171314215 hasConcept C41065033 @default.
• W3171314215 hasConcept C52102323 @default.
• W3171314215 hasConcept C54355233 @default.

• next