AI Paper Atlas

Research map / Image generation / GANs

Dynamic Facial Expression Generation on Hilbert Hypersphere With Conditional Wasserstein Generative Adversarial Nets

Naima Otberdout, Mohamed Daoudi, Anis Kacem, Lahoucine Ballihi, Stefano Berretti

TPAMI 2020 · Image generation › GANs · approach: GAN-based image manipulation

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add identity loss L_id to TextureGAN (ablation)full model · Oulu-CASIAACD: 0.0193 -> 0.016Improved
add identity loss L_id to TextureGAN (ablation)full model · Oulu-CASIAPSNR: 26.05 -> 24.443Hurt
add reconstruction loss on S to MotionGAN (ablation)full model · Oulu-CASIAPSNR: 24.180 -> 24.443Improved
add reconstruction loss on S to MotionGAN (ablation)full model · Oulu-CASIASSIM: 0.886 -> 0.891Improved
add reconstruction loss on S to MotionGAN (ablation)full model · Oulu-CASIAACD: 0.017 -> 0.016Improved
use ground truth landmarks instead of MotionGAN-generated ones (ablation)full model · Oulu-CASIAACD-I: 0.107 -> 0.10Improved
use ground truth landmarks instead of MotionGAN-generated ones (ablation)full model · Oulu-CASIAPSNR: 24.443 -> 25.98Improved
generate with MotionGAN+TextureGAN instead of VGANvideo generation · MUGInception Score: 3.24 -> 3.52Improved
generate with MotionGAN+TextureGAN instead of VGANvideo generation · MUGPSNR: 20.98 -> 25.90Improved
generate with MotionGAN+TextureGAN instead of MoCoGANvideo generation · MUGACD: 0.045 -> 0.011Improved
generate with ours instead of VGAN, LSTM recognitionvideo generation · MUG expression recognitionaccuracy (%): 45.28 -> 69.82Improved
augment training data with MotionGAN videos (x10)LSTM expression classifier · CASIAaccuracy (%): 87.5 -> 92.7Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy