Research on Offline Handwritten Signature Verification explored a large variety of handcrafted feature extractors, ranging from graphology, texture descriptors to interest points. In spite of advancements in the last decades, performance of such systems is still far from optimal when we test the systems against skilled forgeries - signature forgeries that target a particular individual. In previous research, we proposed a formulation of the problem to learn features from data (signature images) in a Writer-Independent format, using Deep Convolutional Neural Networks (CNNs), seeking to improve performance on the task. In this research, we push further the performance of such method, exploring a range of architectures, and obtaining a large improvement in state-of-the-art performance on the GPDS dataset, the largest publicly available dataset on the task. In the GPDS-160 dataset, we obtained an Equal Error Rate of 2.74%, compared to 6.97% in the best result published in literature (that used a combination of multiple classifiers). We also present a visual analysis of the feature space learned by the model, and an analysis of the errors made by the classifier. Our analysis shows that the model is very effective in separating signatures that have a different global appearance, while being particularly vulnerable to forgeries that very closely resemble genuine signatures, even if their line quality is bad, which is the case of slowly-traced forgeries.

DeepMind has built a neural network that can access an external memory like a conventional Turing machine. The result is a computer that mimics the short-term memory of the human brain.

Deep neural networks that learn to represent data in multiple layers of increasing abstraction have dramatically improved the state-of-the-art for speech recognition, object recognition, object detection, predicting the activity of drug molecules, and many other tasks. Deep learning discovers intricate structure in large datasets by building distributed representations, either via supervised, unsupervised or reinforcement learning.

Apple wants you to know that Siri is no dunderhead. The normally secretive tech giant lifted the lid on its artificial intelligence efforts by inviting the writer Steven Levy to its headquarters to hear it crow about all the products and services that are benefiting from various machine-learning techniques, especially deep learning. The resulting article, published on Levy's site, Backchannel, paints a picture of a company keen to boast about its AI chops. It's certainly interesting to hear about how deep learning is being used to improve Siri's hearing, voice, and natural language; and how machine learning is helping to make other Apple products smarter. But it's also hard not to notice that the company is starting to lag behind some of its competitors.

This is a blog series in several parts -- where I describe my experiences and go deep into the reasons behind my choices. In Part 1, I discussed the pros and cons of different symbolic frameworks, and my reasons for choosing Theano (with Lasagne) as my platform of choice. Part 2 of this blog series is based on my upcoming talk at The Data Science Conference, 2016. Here in Part 2, I describe Deep Convolutional Neural Networks (DCNNs) and how Transfer learning and Fine-tuning helps better the training process for domain specific images. Please feel free to email me at [email protected] if you have questions.

TensorFlow is a framework created by Google for creating Deep Learning models. Deep Learning is a category of machine learning models that use multi-layer neural networks. The idea of deep learning has been around since 1943 when neurophysiologist Warren McCulloch and mathematician Walter Pitts wrote a paper on how neurons might work and they modeled a simple neural network using electrical circuits. Many, many developments have occurred since then. These highly accurate mathematical models are extremely computationally expensive.

Three years earlier, Apple had been the first major tech company to integrate a smart assistant into its operating system. Siri was the company's adaptation of a standalone app it had purchased, along with the team that created it, in 2010. Initial reviews were ecstatic, but over the next few months and years, users became impatient with its shortcomings. All too often, it erroneously interpreted commands. So Apple moved Siri voice recognition to a neural-net based system for US users on that late July day (it went worldwide on August 15, 2014.)

There are so many articles about learning Deep Learning but still I decided to write one more. The reason is I find many of those articles saying the same thing over and over again. I think there is a need for a new guide for learning DL for people who are already well-versed with traditional ML. Deep Learning is as much science as it is art. I've met and spoken to a lot of people recently who believe doing deep learning is pretty easy, you only need an open source library like TensorFlow, Theano etc. and decent data at your disposal, and you are all set.

Prisma, the viral photography app that retouches your pics to resemble the work of famous artists, has updated its iOS app to allow for offline image processing. Half of the company's filters (that's 16 in total) are now available offline, with more in the pipeline, according to Prisma. In the past, the app would transform the look of your images into the style of a famous artist (such as Pablo Picasso or Roy Lichtenstein) by feeding the photo through an artificial intelligence. The AI would then reinterpret your image using a deep learning method known as neural networks. With its latest update for iOS, that algorithm will be available offline on a smartphone for the first time.

The purpose of this book is to help to spread TensorFlow knowledge among engineers who want to expand their wisdom in the exciting world of Machine Learning. We believe that anyone with an engineering background might require from now on Deep Learning, and Machine Learning in general, to apply it in their work. As the title indicates, it is a first contact with TensorFlow in order to get started with Deep Learning programming. The book has a practical nature, and therefore it reduces the theoretical part as much as possible, assuming that the reader has some basic understanding about Machine Learning.