There has been much recent talk about the near future of code writing itself with the help of trained neural networks but outside of some limited use cases, that reality is still quite some time away--at least for ordinary development efforts. Although auto-code generation is not a new concept, it has been getting fresh attention due to better capabilities and ease of use in neural network frameworks. But just as in other areas where AI is touted as being the near-term automation savior, the hype does not match the technological complexity need to make it reality. Just in the last few weeks Google, Microsoft and IBM have announced new ways of boosting developer productivity with deep learning frameworks that fill themselves in--at least in part. The headlines exclaim that code is writing itself; that programmers will no longer be necessary.

In one of our recent articles we told about the advances in Deep Learning text and speech applications. An equally prominent domain is the DL algorithms for machine perception. Google Brain team has developed a new Optical Character Recognition (OCR) algorithm to improve Google Maps and Street View services with better road sign recognition capabilities. Due to this they were able to make the road signs better visible on more than 80 billion photographs. The resulting model is also able to concentrate on the shop signs automatically, filtering them out of the unneeded visual noise on the photos.

When looking for another subject to learn, my ideas went around various AI problems. I wrote something about Deep Neural Networks in Swift (here and here) last year and I want to continue doing so. I made up myself learning plan which was structured like Deep Learning Artificial Intelligence basics Reinforcement Learning. But while waiting on the 5th part of Andrew's Ng Deep Learning MOOC I found about the Reinforced Learning: An Introduction (which I might refer as the book in those posts) and got hooked. What this series will be my journey through this very exciting area.

Artificial intelligence is taking the automobile industry by storm while all the major automobile players are utilizing their resources and technology to come up with the best. The beauty of devices with artificial intelligence is that it tries to learn from sensory inputs like real sounds and images. In the same way, when intelligence is applied to the technology within an automobile, it would recognize the environment and evaluate the contextual implications when it moves or faces any hurdles. In 2015, the install rate of AI based systems in new vehicles was just 8%; this number is expected to soar to 109% in 2025. This is because different kinds of AI systems will be installed in vehicles.

The revolution of the web has led to an explosion of content generated every day on the internet. Social sharing platforms such as Facebook, Twitter, Instagram, etc. have seen astonishing growth in their daily active users, but have been at their split ends when it comes to monitoring the content generated by users. Users are uploading inappropriate content such as nudity or using abusive language while commenting on posts. Such behavior leads to social issues like bullying and revenge porn and also hampers the authenticity of the platform. However, the pace at which the content is generated online today is so high that it is nearly impossible to monitor everything manually.

The field of medicine has, arguably, been more positively affected by modern deep learning techniques than any other industry. And, despite the unending deluge of panic-ridden articles declaring AI the path to apocalypse, we're now living in a world where algorithms save lives every day. Thanks to AI, an iPhone can detect cancer and a smart watch can detect a stroke. Machine learning is infiltrating and optimizing nearly every aspect of medicine from the way 911 emergency services are dispatched to assisting doctors during surgery. You can even quit smoking or kick your opiate addiction with the help of AI. Emergency service dispatchers in Copenhagen are using a new voice-assistant technology called Corti.

X-ray computed tomography (CT) using sparse projection views is a recent approach to reduce the radiation dose. However, due to the insufficient projection views, an analytic reconstruction approach using the filtered back projection (FBP) produces severe streaking artifacts. Recently, deep learning approaches using large receptive field neural networks such as U-Net have demonstrated impressive performance for sparse- view CT reconstruction. However, theoretical justification is still lacking. Inspired by the recent theory of deep convolutional framelets, the main goal of this paper is, therefore, to reveal the limitation of U-Net and propose new multi-resolution deep learning schemes. In particular, we show that the alternative U- Net variants such as dual frame and the tight frame U-Nets satisfy the so-called frame condition which make them better for effective recovery of high frequency edges in sparse view- CT. Using extensive experiments with real patient data set, we demonstrate that the new network architectures provide better reconstruction performance.

In sequence-based generative models, besides the generation of samples likely to have been drawn from a data distribution, it is often desirable to finetune the samples towards some domain-specific metrics. This work proposes a method to guide the structure and quality of samples utilizing a combination of adversarial training and expert-based rewards with reinforcement learning. Building on SeqGAN, a sequence based Generative Adversarial Network (GAN) framework modeling the data generator as a stochastic policy in a reinforcement learning setting, we extend the training process to include domain-specific objectives additional to the discriminator reward. The mixture of both types of rewards can be controlled via a tune-able parameter. To improve training stability we utilize the Wasserstein distance as loss function for the discriminator. We demonstrate the effectiveness of this approach in two tasks: generation of molecules encoded as text sequences and musical melodies. The experimental results demonstrate the models can generate samples which maintain information originally learned from data, retain sample diversity, and show improvement in the desired metrics.

Bayesian optimization has become a standard technique for hyperparameter optimization, including data-intensive models such as deep neural networks that may take days or weeks to train. We consider the setting where previous optimization runs are available, and we wish to use their results to warm-start a new optimization run. We develop an ensemble model that can incorporate the results of past optimization runs, while avoiding the poor scaling that comes with putting all results into a single Gaussian process model. The ensemble combines models from past runs according to estimates of their generalization performance on the current optimization. Results from a large collection of hyperparameter optimization benchmark problems and from optimization of a production computer vision platform at Facebook show that the ensemble can substantially reduce the time it takes to obtain near-optimal configurations, and is useful for warm-starting expensive searches or running quick re-optimizations.

Estimating the travel time of a path is of great importance to smart urban mobility. Existing approaches are either based on estimating the time cost of each road segment which are not able to capture many cross-segment complex factors, or designed heuristically in a non-learning-based way which fail to utilize the existing abundant temporal labels of the data, i.e., the time stamp of each trajectory point. In this paper, we leverage on new development of deep neural networks and propose a novel auxiliary supervision model, namely DeepTravel, that can automatically and effectively extract different features, as well as make full use of the temporal labels of the trajectory data. We have conducted comprehensive experiments on real datasets to demonstrate the out-performance of DeepTravel over existing approaches.