In the summer of 2018, nearly 70 civil rights and research organizations wrote a letter to Jeff Bezos demanding that Amazon stop providing face recognition technology to governments. As part of an increased focus on the role that tech companies were playing in enabling the US government’s tracking and deportation of immigrants, it called on Amazon to “stand up for civil rights and civil liberties.” “As advertised,” it said, “Rekognition is a powerful surveillance system readily available to violate rights and target communities of color.”
Along with the letter, the American Civil Liberties Union (ACLU) of Washington delivered over 150,000 petition signatures as well as another letter from the company’s own shareholders expressing similar demands. A few days later, Amazon’s employees echoed the concerns in an internal memo.
Despite the mounting pressure, Amazon continued with business as usual. Read More
Daily Archives: June 15, 2020
How to jam neural networks
Deep neural networks (DNNs) have been a very active field of research for eight years now, and for the last five we’ve seen a steady stream of adversarial examples – inputs that will bamboozle a DNN so that it thinks a 30mph speed limit sign is a 60 instead, and even magic spectacles to make a DNN get the wearer’s gender wrong.
So far, these attacks have targeted the integrity or confidentiality of machine-learning systems. Can we do anything about availability? Read More
PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of Generative Models
The primary aim of single-image super-resolution is to construct a high-resolution (HR) image from a corresponding low-resolution (LR) input. In previous approaches,which have generally been supervised, the training objective typically measures a pixel-wise average distance be-tween the super-resolved (SR) and HR images. Optimizing such metrics often leads to blurring, especially in high variance (detailed) regions. We propose an alternative formulation of the super-resolution problem based on creating realistic SR images that downscale correctly. We present a novel super-resolution algorithm addressing this problem, PULSE (Photo Up sampling via Latent Space Exploration), which generates high-resolution, realistic images at resolutions previously unseen in the literature. It accomplishes this in an entirely self-supervised fashion and is not confined to a specific degradation operator used during training, unlike previous methods (which require training on databases of LR-HR image pairs for supervised learning). Instead of starting with the LR image and slowly adding detail, PULSE traverses the high-resolution natural image manifold, searching for images that downscale to the original LR image. This is formalized through the “down-scaling loss,” which guides exploration through the latent space of a generative model. By leveraging properties of high-dimensional Gaussians, we restrict the search space to guarantee that our outputs are realistic. PULSE thereby generates super-resolved images that both are realistic and downscale correctly. We show extensive experimental results demonstrating the efficacy of our approach in the do-main of face super-resolution (also known as face hallucination). Our method outperforms state-of-the-art methods in perceptual quality at higher resolutions and scale factors than previously possible. Read More
Rick's Cafe AI 