The University of Amsterdam is one of Europe’s leading research universities, renowned for its work in artificial intelligence, computer vision, and machine learning. With a strong emphasis on ethical AI, multimodal learning, and data-driven innovation, it plays a central role in advancing responsible and scalable technologies across academia and industry. NEC Laboratories America collaborated with the University of Amsterdam on advancing domain generalization techniques for face anti-spoofing. We jointly introduced a novel method using gradient signal analysis to enhance the robustness of face authentication systems across various datasets, thereby improving their reliability in diverse environments.

Posts

Channel-Recurrent Autoencoding for Image Modeling

/in /by

Despite recent successes in synthesizing faces and bedrooms, existing generative models struggle to capture more complex image types (Figure 1), potentially due to the oversimplification of their latent space constructions. To tackle this issue, building on Variational Autoencoders (VAEs), we integrate recurrent connections across channels to both inference and generation steps, allowing the high-level features to be captured in global-to-local, coarse-to-fine manners. Combined with adversarial loss, our channel-recurrent VAE-GAN (crVAE-GAN) outperforms VAE-GAN in generating a diverse spectrum of high resolution images while maintaining the same level of computational efficacy. Our model produces interpretable and expressive latent representations to benefit downstream tasks such as image completion. Moreover, we propose two novel regularizations, namely the KL objective weighting scheme over time steps and mutual information maximization between transformed latent variables and the outputs, to enhance the training.

Read more