If you search "Marketing and AI", Google returns almost four million results. Recent headlines range from the strategic - AI Will Make Marketing Less Manual (well, duh) – to the tactical, like 4 ways AI can improve email marketing. The hype machine is in full swing. This is a hot topic at my company, Conversion Logic, where we use machine learning algorithms for marketing attribution. It's tempting, from our own marketing perspective, to jump on the bandwagon and "AI-wash" our offering.

Computer scientists are developing a mind-reading computer that deciphers symbols that people have looked at. The device accurately replicates shapes seen. The computer scans brain activity, then successfully redraws those numerals and symbols, say scientists working on the project. It's a "step towards a direct'telepathic' connection between brains and computers," said the Chinese Academy of Sciences (CAS) in a May news article. And indeed, should it work reliably, it would be a significant improvement on simple Functional Magnetic Resonance Imaging (fMRI) scans, which just read activity in parts of the brain and are used primarily for research.

Given a particular input, one can often predict how a person will respond. That is not the case for the most intelligent machines in our midst. The creators of AlphaGo -- a computer program built by Google's DeepMind that decisively beat the world's finest human player of the board game Go -- admitted they could not have divined its winning moves. This unpredictability, also seen in the Facebook chatbots that were shut down after developing their own language, has stirred disquiet in the field of artificial intelligence.

Start computer vision research using deep learning much easier ChainerCV Latest algorithms with your data Provide complete model code, training code, inference code for segmentation algorithms (SegNet, etc.) and object detection algorithms (Faster R-CNN, SSD, etc.), and so on All code is confirmed to reproduce the results All training code and model code reproduced the experimental results shown in the original paper https://github.com/pfnet/chainercv

With AI and big data analytics powering the next generation of surgical robots, 3 most promising AI systems can be incorporated in future surgical robots: IBM Watson, Alpha Go, and machine learning algorithms. Watson is capable of becoming an intelligent surgical assistant capable of storing a plethora of medical information and use natural language processing to respond to surgeon's queries. Google's DeepMind project AlphaGo can be a potential contender for surgical robotic AI systems. Moreover, unsupervised pattern matching algorithms would aid doctors in recognizing when a sequence of symptoms results in a particular disease.

Computer Vision generates mathematical models from images; Computer Graphics draws in images from models and lastly image processing takes image as an input and gives an image at the output. Computer Vision is an overlapping field drawing on concepts from areas such as artificial intelligence, digital image processing, machine learning, deep learning, pattern recognition, probabilistic graphical models, scientific computing and a lot of mathematics. Watch these videos and alongside implementing the learned concepts and algorithms by following GaTech Prof. James Hays' projects of his Computer Vision class. Have a quick go through Building Machine Learning Systems with Python and Python Machine Learning.

Semantic segmentation was seen as a challenging computer vision problem few years ago. Due to recent advancements in deep learning, relatively accurate solutions are now possible for its use in automated driving. In this paper, the semantic segmentation problem is explored from the perspective of automated driving. Most of the current semantic segmentation algorithms are designed for generic images and do not incorporate prior structure and end goal for automated driving. First, the paper begins with a generic taxonomic survey of semantic segmentation algorithms and then discusses how it fits in the context of automated driving. Second, the particular challenges of deploying it into a safety system which needs high level of accuracy and robustness are listed. Third, different alternatives instead of using an independent semantic segmentation module are explored. Finally, an empirical evaluation of various semantic segmentation architectures was performed on CamVid dataset in terms of accuracy and speed. This paper is a preliminary shorter version of a more detailed survey which is work in progress.

Convolutional Neural Networks have been a subject of great importance over the past decade and great strides have been made in their utility for producing state of the art performance in many computer vision problems. However, the behavior of deep networks is yet to be fully understood and is still an active area of research. In this work, we present an intriguing behavior: pre-trained CNNs can be made to improve their predictions by structurally perturbing the input. We observe that these perturbations - referred as Guided Perturbations - enable a trained network to improve its prediction performance without any learning or change in network weights. We perform various ablative experiments to understand how these perturbations affect the local context and feature representations. Furthermore, we demonstrate that this idea can improve performance of several existing approaches on semantic segmentation and scene labeling tasks on the PASCAL VOC dataset and supervised classification tasks on MNIST and CIFAR10 datasets.

The goal of Machine Learning to automatically learn from data, extract knowledge and to make decisions without any human intervention. Such automatic (aML) approaches show impressive success. Recent results even demonstrate intriguingly that deep learning applied for automatic classification of skin lesions is on par with the performance of dermatologists, yet outperforms the average. As human perception is inherently limited, such approaches can discover patterns, e.g. that two objects are similar, in arbitrarily high-dimensional spaces what no human is able to do. Humans can deal only with limited amounts of data, whilst big data is beneficial for aML; however, in health informatics, we are often confronted with a small number of data sets, where aML suffer of insufficient training samples and many problems are computationally hard. Here, interactive machine learning (iML) may be of help, where a human-in-the-loop contributes to reduce the complexity of NP-hard problems. A further motivation for iML is that standard black-box approaches lack transparency, hence do not foster trust and acceptance of ML among end-users. Rising legal and privacy aspects, e.g. with the new European General Data Protection Regulations, make black-box approaches difficult to use, because they often are not able to explain why a decision has been made. In this paper, we present some experiments to demonstrate the effectiveness of the human-in-the-loop approach, particularly in opening the black-box to a glass-box and thus enabling a human directly to interact with an learning algorithm. We selected the Ant Colony Optimization framework, and applied it on the Traveling Salesman Problem, which is a good example, due to its relevance for health informatics, e.g. for the study of protein folding. From studies of how humans extract so much from so little data, fundamental ML-research also may benefit.

Dropout is a very effective way of regularizing neural networks. Stochastically "dropping out" units with a certain probability discourages over-specific co-adaptations of feature detectors, preventing overfitting and improving network generalization. Besides, Dropout can be interpreted as an approximate model aggregation technique, where an exponential number of smaller networks are averaged in order to get a more powerful ensemble. In this paper, we show that using a fixed dropout probability during training is a suboptimal choice. We thus propose a time scheduling for the probability of retaining neurons in the network. This induces an adaptive regularization scheme that smoothly increases the difficulty of the optimization problem. This idea of "starting easy" and adaptively increasing the difficulty of the learning problem has its roots in curriculum learning and allows one to train better models. Indeed, we prove that our optimization strategy implements a very general curriculum scheme, by gradually adding noise to both the input and intermediate feature representations within the network architecture. Experiments on seven image classification datasets and different network architectures show that our method, named Curriculum Dropout, frequently yields to better generalization and, at worst, performs just as well as the standard Dropout method.