Matches in SemOpenAlex for { <https://semopenalex.org/work/W3047256485> ?p ?o ?g. }
- W3047256485 abstract "We present a deployment friendly, fast bottom-up framework for multi-person 3D human pose estimation. We adopt a novel neural representation of multi-person 3D pose which unifies the position of person instances with their corresponding 3D pose representation. This is realized by learning a generative pose embedding which not only ensures plausible 3D pose predictions, but also eliminates the usual keypoint grouping operation as employed in prior bottom-up approaches. Further, we propose a practical deployment paradigm where paired 2D or 3D pose annotations are unavailable. In the absence of any paired supervision, we leverage a frozen network, as a teacher model, which is trained on an auxiliary task of multi-person 2D pose estimation. We cast the learning as a cross-modal alignment problem and propose training objectives to realize a shared latent space between two diverse modalities. We aim to enhance the model's ability to perform beyond the limiting teacher network by enriching the latent-to-3D pose mapping using artificially synthesized multi-person 3D scene samples. Our approach not only generalizes to in-the-wild images, but also yields a superior trade-off between speed and performance, compared to prior top-down approaches. Our approach also yields state-of-the-art multi-person 3D pose estimation performance among the bottom-up approaches under consistent supervision levels." @default.
- W3047256485 created "2020-08-10" @default.
- W3047256485 creator A5009355682 @default.
- W3047256485 creator A5015371248 @default.
- W3047256485 creator A5062096014 @default.
- W3047256485 creator A5074537472 @default.
- W3047256485 creator A5086834875 @default.
- W3047256485 date "2020-08-04" @default.
- W3047256485 modified "2023-09-26" @default.
- W3047256485 title "Unsupervised Cross-Modal Alignment for Multi-Person 3D Pose Estimation" @default.
- W3047256485 cites W1861492603 @default.
- W3047256485 cites W1943191679 @default.
- W3047256485 cites W1950788856 @default.
- W3047256485 cites W1975868529 @default.
- W3047256485 cites W2101032778 @default.
- W3047256485 cites W2159291411 @default.
- W3047256485 cites W2175012183 @default.
- W3047256485 cites W2215643317 @default.
- W3047256485 cites W2259801182 @default.
- W3047256485 cites W2307770531 @default.
- W3047256485 cites W2382036597 @default.
- W3047256485 cites W2473930607 @default.
- W3047256485 cites W2522527348 @default.
- W3047256485 cites W2554247908 @default.
- W3047256485 cites W2555751471 @default.
- W3047256485 cites W2559085405 @default.
- W3047256485 cites W2604375920 @default.
- W3047256485 cites W2605102758 @default.
- W3047256485 cites W2605947573 @default.
- W3047256485 cites W2611932403 @default.
- W3047256485 cites W2612706635 @default.
- W3047256485 cites W2756050327 @default.
- W3047256485 cites W2766966408 @default.
- W3047256485 cites W2769331938 @default.
- W3047256485 cites W2776685160 @default.
- W3047256485 cites W2788865504 @default.
- W3047256485 cites W2792747672 @default.
- W3047256485 cites W2797184202 @default.
- W3047256485 cites W2804648901 @default.
- W3047256485 cites W2819476901 @default.
- W3047256485 cites W2886970656 @default.
- W3047256485 cites W2888934629 @default.
- W3047256485 cites W2889600678 @default.
- W3047256485 cites W2895748257 @default.
- W3047256485 cites W2901678097 @default.
- W3047256485 cites W2902149428 @default.
- W3047256485 cites W2902214002 @default.
- W3047256485 cites W2945528222 @default.
- W3047256485 cites W2951874610 @default.
- W3047256485 cites W2953216610 @default.
- W3047256485 cites W2963013806 @default.
- W3047256485 cites W2963037989 @default.
- W3047256485 cites W2963207848 @default.
- W3047256485 cites W2963402313 @default.
- W3047256485 cites W2963420272 @default.
- W3047256485 cites W2963598138 @default.
- W3047256485 cites W2963995996 @default.
- W3047256485 cites W2964121744 @default.
- W3047256485 cites W2968529893 @default.
- W3047256485 cites W2968940310 @default.
- W3047256485 cites W2969696852 @default.
- W3047256485 cites W2975775043 @default.
- W3047256485 cites W2981801509 @default.
- W3047256485 cites W2991631290 @default.
- W3047256485 cites W2993728126 @default.
- W3047256485 cites W2997288107 @default.
- W3047256485 cites W2998273692 @default.
- W3047256485 cites W3009839269 @default.
- W3047256485 cites W3035072447 @default.
- W3047256485 cites W3037367276 @default.
- W3047256485 cites W3098473649 @default.
- W3047256485 cites W3101442004 @default.
- W3047256485 hasPublicationYear "2020" @default.
- W3047256485 type Work @default.
- W3047256485 sameAs 3047256485 @default.
- W3047256485 citedByCount "1" @default.
- W3047256485 countsByYear W30472564852021 @default.
- W3047256485 crossrefType "posted-content" @default.
- W3047256485 hasAuthorship W3047256485A5009355682 @default.
- W3047256485 hasAuthorship W3047256485A5015371248 @default.
- W3047256485 hasAuthorship W3047256485A5062096014 @default.
- W3047256485 hasAuthorship W3047256485A5074537472 @default.
- W3047256485 hasAuthorship W3047256485A5086834875 @default.
- W3047256485 hasConcept C105339364 @default.
- W3047256485 hasConcept C111919701 @default.
- W3047256485 hasConcept C119857082 @default.
- W3047256485 hasConcept C127413603 @default.
- W3047256485 hasConcept C153083717 @default.
- W3047256485 hasConcept C153180895 @default.
- W3047256485 hasConcept C154945302 @default.
- W3047256485 hasConcept C17744445 @default.
- W3047256485 hasConcept C185592680 @default.
- W3047256485 hasConcept C188027245 @default.
- W3047256485 hasConcept C199539241 @default.
- W3047256485 hasConcept C201995342 @default.
- W3047256485 hasConcept C2776359362 @default.
- W3047256485 hasConcept C2780451532 @default.
- W3047256485 hasConcept C31972630 @default.
- W3047256485 hasConcept C36613465 @default.
- W3047256485 hasConcept C41008148 @default.