Our brains are wired in a way that they can differentiate between objects, both living and non-living by simply looking at them. In fact, the recognition of objects and a situation through visualization is the fastest way to gather, as well as to relate information. This becomes a pretty big deal for computers where a vast amount of data has to be stuffed into it, before the computer can perform an operation on its own. Ironically, with each passing day, it is becoming essential for machines to identify objects through facial recognition, so that humans can take the next big step towards a more scientifically advanced social mechanism. So, what progress have we really made in that respect?

Last month an image purporting to show children in cages as a result of current immigration policies went viral on social media, accelerated by a number of high profile journalists, activists and former government officials who shared it widely – their visibility and stature leading many to trust the image at face value without the level of suspicion and verification that users might apply to other viral images. The image was real, but taken out of context and spread virally before users began to realize it actually dated from a 2014 news article. Yet, when I first saw the image I simply right-clicked on it and ran a reverse Google Images search that immediately turned up the original 2014 source. Could social media outlets like Twitter and Facebook automate such image searches to help combat fake news at scale? Social media today is an ocean of false and misleading information spread for nefarious purposes, but far more often by well-meaning individuals who share first and ask questions later.

Melbourne man Milorad "Michael" Trkulja has won his high court battle to sue the search engine Google for defamation over images and search results that link him to the Melbourne criminal underworld. Trkulja said he would continue legal action against Google until it removed his name and photos from the internet. Trkulja, who was shot in the back in a Melbourne restaurant in 2004, successfully argued in the Victorian supreme court in 2012 that Google defamed him by publishing photos of him linked to hardened criminals of Melbourne's underworld. Four years later the Victorian court of appeal overturned the decision, finding the case had no prospect of successfully proving defamation. The high court disputed that ruling in a judgment on Wednesday and ordered Google to pay Trkulja's legal costs.

The two major tasks in medical imaging that appear to be naturally predestined to be solved with AI algorithms are segmentation and classification. Most of techniques used in medical imaging were conventional image processing, or more widely formulated computer vision algorithms. One can find many works with artificial neural networks, the backbone of deep learning. However, most works were focused on conventional computer vision which focused, and still does, on "handcrafted" features, techniques that were the results of manual design to extract useful and differentiating information from medical images. Some progress was visible in the late 90s and early 2000s (for instance, the SIFT method in 1999, or visual dictionaries in early 2000s) but there were no breakthroughs.

The two major tasks in medical imaging that appear to be naturally predestined to be solved with AI algorithms are segmentation and classification. Most of techniques used in medical imaging were conventional image processing, or more widely formulated computer vision algorithms. One can find many works with artificial neural networks, the backbone of deep learning. However, most works were focused on conventional computer vision which focused, and still does, on "handcrafted" features, techniques that were the results of manual design to extract useful and differentiating information from medical images. Some progress was visible in the late 90s and early 2000s (for instance, the SIFT method in 1999, or visual dictionaries in early 2000s) but there were no breakthroughs.

About 10 years ago, it would have been hard to believe that you could ask a Bluetooth speaker for a classic cheese soufflé recipe or take a picture of an object using your phone and find out exactly where to purchase it. These interactions have been primarily realized through advancements in machine learning AI. One of the biggest developments in AI over the past three years has been in the area of voice recognition and natural language processing and we're starting to see advancements in more complex human machine interaction in the form of image/video search. Forward-thinking businesses are already using this new form of machine learning AI image recognition to allow users to search for products using pictures to find the same or similar looks and outfits they stock. However, does this mean intelligent image search is the next big thing? Major search engines have supported a form of'image search' for some time.

In the 2012 Hollywood movie "Act of Valor", US Navy Seals launch a Raven UAV to procure live video streaming for identifying targets prior to a raid. In yet another movie, "Zero Dark Thirty", analysts use FMV/ drone feed to re-create Bin Laden's compound for training US Navy Seals, and later guide them to accomplish their mission. The growing use of image analytics in tracking, detecting, analyzing and predicting outcomes have been effectively used by story writers to piece amazing stories. Image analytics is a technique by which an image is digitally processed for extracting and analyzing data for insightful information. Big data still remains a scary and invincible concept, because of the unmanageable amount of unstructured data present in it.

Mitek Systems, Inc., a leader in digital identity verification solutions, announced that it has acquired A2iA, a leader in artificial intelligence (AI) and image analysis. The deal is for €42.5 million in cash and shares of Mitek's common stock. Mitek software is used in 6,100 U.S. banks, including all of the top 10 largest U.S. financial institutions. "The acquisition of A2iA combines two market leaders in image recognition and processing, creating a powerful force with a deep expertise in image analytics," industry expert Bob Meara, senior analyst at Celent said in a press release. A2iA uses AI and machine learning to create proprietary algorithms that process millions of checks, IDs and documents each day for banks, retailers, insurance companies, mobile operators, healthcare providers and governments in more than 42 countries and 11 languages.

We describe an adversarial learning approach to constrain convolutional neural network training for image registration, replacing heuristic smoothness measures of displacement fields often used in these tasks. Using minimally-invasive prostate cancer intervention as an example application, we demonstrate the feasibility of utilizing biomechanical simulations to regularize a weakly-supervised anatomical-label-driven registration network for aligning pre-procedural magnetic resonance (MR) and 3D intra-procedural transrectal ultrasound (TRUS) images. A discriminator network is optimized to distinguish the registration-predicted displacement fields from the motion data simulated by finite element analysis. During training, the registration network simultaneously aims to maximize similarity between anatomical labels that drives image alignment and to minimize an adversarial generator loss that measures divergence between the predicted- and simulated deformation. The end-to-end trained network enables efficient and fully-automated registration that only requires an MR and TRUS image pair as input, without anatomical labels or simulated data during inference. 108 pairs of labelled MR and TRUS images from 76 prostate cancer patients and 71,500 nonlinear finite-element simulations from 143 different patients were used for this study. We show that, with only gland segmentation as training labels, the proposed method can help predict physically plausible deformation without any other smoothness penalty. Based on cross-validation experiments using 834 pairs of independent validation landmarks, the proposed adversarial-regularized registration achieved a target registration error of 6.3 mm that is significantly lower than those from several other regularization methods.

News Blog China's Perfect Storm AI Moment May 22, 2018 Robin Raskin, Founder, Living in Digital Times In July of 2017 the Chinese government issued a development plan to make the country the world leader in artificial intelligence by 2030. So far things are going better than pretty well. The government's commitment to AI dominance, the sheer amount of data that a country as large as China can feed to its AI learning systems, a mobile-dominated infrastructure, and an edge on chip manufacturing for AI-intensive activities like facial and image recognition contribute to China's AI success story. Deep machine learning, the ability for machines to ingest data, learn from it and then make predictions from it requires lots of data. "There are 160 cities in China with over 1 million people; there are 10 in the U.S., says Deborah Weinswig, Senior Analyst for Fung.