Google Brain has been hard at work making the future happen now. Its latest effort brings a neural network approach to increasing the resolution of images. The results are like magic:
The middle column is the predicted high resolution image. To see that there is actually more information in the 8x8 image try looking at them through half closed eyes.
We present a pixel recursive super resolution model that synthesizes realistic details into images while enhancing their resolution. A low resolution image may correspond to multiple plausible high resolution images, thus modeling the super resolution process with a pixel independent conditional model often results in averaging different details--hence blurry edges.
By contrast, our model is able to represent a multimodal conditional distribution by properly modeling the statistical dependencies among the high resolution image pixels, conditioned on a low resolution input. We employ a PixelCNN architecture to define a strong prior over natural images and jointly optimize this prior with a deep conditioning convolutional network. Human evaluations indicate that samples from our proposed model look more photo realistic than a strong L2 regression baseline.
It has been recently shown that Generative Adversarial Networks (GANs) can produce synthetic images of exceptional visual fidelity. In this work, we propose the GAN-based method for automatic face aging. Contrary to previous works employing GANs for altering of facial attributes, we make a particular emphasize on preserving the original person's identity in the aged version of his/her face. To this end, we introduce a novel approach for "Identity-Preserving" optimization of GAN's latent vectors. The objective evaluation of the resulting aged and rejuvenated face images by the state-of-the-art face recognition and age estimation solutions demonstrate the high potential of the proposed method.
It is very easy to fool yourself into thinking that some clever AI technique is working well when in fact it is really just picking up accidental and irrelevant features. Some years ago one of the first uses of single layer perceptrons was to tell males from females in photos. It managed an impressive accuracy, but on further analysis it turned out the it was simply reacting to the size of the dark area at the top of each normalized photograph, i.e. it was determining gender on the basis of amount of hair. It seems that something similar might be going on with the deduction of gender from iris patterns.
Predicting a person's gender based on the iris texture has been explored by several researchers. This paper considers several dimensions of experimental work on this problem, including person-disjoint train and test, and the effect of cosmetics on eyelash occlusion and imperfect segmentation.
We also consider the use of multi-layer perceptron and convolutional neural networks as classifiers, comparing the use of data-driven and hand-crafted features.
Our results suggest that the gender-from-iris problem is more difficult than has so far been appreciated. Estimating accuracy using a mean of N person-disjoint train and test partitions, and considering the effect of makeup - a combination of experimental conditions not present in any previous work - we find a much weaker ability to predict gender-from-iris texture than has been suggested in previous work.
Microsoft has announced the General Availability on Azure of its Face API and Computer Vision API to give developers easy tools for handling and interpreting images. The Content Moderator is the third [ ... ]