A supporter of Maduro holds a painting of him in Caracas. A supporter of Maduro holds a painting of him in Caracas. Minutes after Donald Trump announced a "large-scale strike" against Venezuela early on Saturday morning, false and misleading AI-generated images began flooding social media. There were fake photos of Nicolás Maduro being escorted off a plane by US law enforcement agents, images of jubilant Venezuelans pouring into the streets of Caracas and videos of missiles raining down on the city - all fake. The fabricated content intermixed with real videos and photos of US aircraft flying over the Venezuelan capital and explosions lighting up the dark sky.

Researchers often spend weeks sifting through decades of unlabeled satellite imagery(on NASA Worldview) in order to develop datasets on which they can start conducting research. We developed an interactive, scalable and fast image similarity search engine (which can take one or more images as the query image) that automatically sifts through the unlabeled dataset reducing dataset generation time from weeks to minutes. In this work, we describe key components of the end to end pipeline. Our similarity search system was created to be able to identify similar images from a potentially petabyte scale database that are similar to an input image, and for this we had to break down each query image into its features, which were generated by a classification layer stripped CNN trained in a supervised manner. To store and search these features efficiently, we had to make several scalability improvements. To improve the speed, reduce the storage, and shrink memory requirements for embedding search, we add a fully connected layer to our CNN make all images into a 128 length vector before entering the classification layers. This helped us compress the size of our image features from 2048 (for ResNet, which was initially tried as our featurizer) to 128 for our new custom model. Additionally, we utilize existing approximate nearest neighbor search libraries to significantly speed up embedding search. Our system currently searches over our entire database of images at 5 seconds per query on a single virtual machine in the cloud. In the future, we would like to incorporate a SimCLR based featurizing model which could be trained without any labelling by a human (since the classification aspect of the model is irrelevant to this use case).

A reverse image search is a technique that allows finding things, people, brands, etc. using a photo. While performing a regular search you usually type a word or phrase that is related to the information you are trying to find; when you do a reverse image search, you upload a picture to a search engine. In the results of regular searches, you receive a list of websites that are connected to these phrases. When you perform a reverse image search, in the results you receive photos of similar things, people, etc, linked to websites about them. Reverse search by image is the best solution to use when looking for similar images, smaller/bigger versions of them, or twin content.

A new tool is promising to make your pictures undetectable to facial recognition software without significantly changing their appearance. Known as Photo Ninja, the tool uses artificial intelligence (AI) to make minor alterations to photos that reportedly confuse facial recognition algorithms. Developed by DoNotPay, the company behind the app that uses AI to provide legal services, Photo Ninja moves the location of certain pixels, alters colors, and adds unnoticeable objects into images. "Photo Ninja uses a novel series of steganography, detection perturbation, visible overlay, and several other AI-based enhancement processes to shield your images from reverse image searches without compromising the look of your photo," the company said. CEO Joshua Browder revealed on Twitter Monday that Photo Ninja will be integrated into the DoNotPay app for paying customers.

A network of fictional journalists, analysts, and political consultants has been used to place opinion pieces favorable to certain Gulf states in a range of media outlets, an investigation from The Daily Beast has revealed. At least 19 fake personas were used to author op-eds published in dozens of mainly conservative publications, with AI-generated headshots of would-be authors used to trick targets into believing the writers were real people. It's not the first time AI has been used in this way, though it's unusual to see machine learning tech deployed for online misinformation in the wild. Last year, a report from The Associated Press found a fake profile on LinkedIn, part of a network of likely spies trying to make connections with professional targets, that also used an AI-generated headshot. AI-generated profile pictures created by sites like ThisPersonDoesNotExist.com have some unique advantages when it comes to building fake online personas.

In this code pattern, work through the process of analyzing an image data set using a pre-trained convolution network (VGG16) and extracting feature vectors for each image using a Jupyter Notebook. Machine learning algorithms provide many useful tools that solve real-world problems. One of the domains that machine learning has had great success with is image recognition. By using computational power to identify images and compare them to other images, you can use machines to perform tasks that a few years ago could be done only by humans. Engineers and data scientists who work with image recognition can encounter a few challenges that can put limits on the work that can be done with machine learning algorithms.

So is DuckDuckGo no good? Surprised you did not mention it. Following last week's article about privacy and surveillance capitalism, several readers wrote in about the absence of DuckDuckGo, and it was mentioned a dozen times in the comments. I have suggested this privacy-oriented search engine a few times since 2012, and I think it's worth a go. However, I'm answering Murray's earlier query along the same lines because I can use his email verbatim rather than cobbling together a joint question from multiple sources.

The recent explosion of false claims in social media and on the Web in general has given rise to a lot of manual fact-checking initiatives. Unfortunately, the number of claims that need to be fact-checked is several orders of magnitude larger than what humans can handle manually. Thus, there has been a lot of research aiming at automating the process. Interestingly, previous work has largely ignored the growing number of claims about images. This is despite the fact that visual imagery is more influential than text and naturally appears alongside fake news. Here we aim at bridging this gap. In particular, we create a new dataset for this problem, and we explore a variety of features modeling the claim, the image, and the relationship between the claim and the image. The evaluation results show sizable improvements over the baseline. We release our dataset, hoping to enable further research on fact-checking claims about images.

The application uses machine learning and, by extension, artificial intelligence, to properly identify people's faces and surface relevant factoids. Users can access Who's Who either through Sky News's mobile app or its website. The idea is to provide digital onlookers second-screen content to fill in gaps during the event. The app will, in real time, identify the faces of people at the wedding. According to a press release, the app uses Amazon Rekognition tools to name people in the crowd and then surface biographical information about them. As Sky News explains it, Who's Who will "[name] wedding guests as they arrive for the ceremony and tells people about their connection to the royal couple."

I've been making some TensorFlow examples for my website, fomoro.com, While it's fresh in my head, I wanted to write up an end-to-end description of what it's like to build a machine learning app, and more specifically, how to make your own reverse image search. For this demo, the work is ⅓ data munging/setup, ⅓ model development and ⅓ app development. At a high-level, I use TensorFlow to create an autoencoder, train it on a bunch of images, use the trained model to find related images, and display them with a Flask app. In the last post, I talked model development and training.