Machine learning is vulnerable to a wide variety of attacks. It is now well understood that by changing the underlying data distribution, an adversary can poison the model trained with it or introduce backdoors. In this paper we present a novel class of training-time attacks that require no changes to the underlying dataset or model Read more about Attacking ML systems by changing the order of the training data[…]
The US Copyright Office and its director Shira Perlmutter have been sued for rejecting one man’s request to register an AI model as the author of an image generated by the software. You guessed correct: Stephen Thaler is back. He said the digital artwork, depicting railway tracks and a tunnel in a wall surrounded by Read more about US Copyright Office sued for denying AI model authorship[…]
We show how a malicious learner can plant an undetectable backdoor into a classifier. On the surface, such a backdoored classifier behaves normally, but in reality, the learner maintains a mechanism for changing the classification of any input, with only a slight perturbation. Importantly, without the appropriate “backdoor key”, the mechanism is hidden and cannot Read more about Planting Undetectable Backdoors in Machine Learning Models[…]
[…] A team from Google, the National University of Singapore, Yale-NUS College, and Oregon State University demonstrated it was possible to extract credit card details from a language model by inserting a hidden sample into the data used to train the system. The attacker needs to know some information about the structure of the dataset, Read more about ML models models leak data after poisoning training data[…]
[…] Known as inverse rendering, the process uses AI to approximate how light behaves in the real world, enabling researchers to reconstruct a 3D scene from a handful of 2D images taken at different angles. The NVIDIA Research team has developed an approach that accomplishes this task almost instantly — making it one of the Read more about NeRF Research Turns a few dozen 2D Photos Into 3D Scenes really quickly[…]
This is an art project by Tijmen Schep that shows how face detection algoritms are increasingly used to judge you. It was made as part of the European Union’s Sherpa research program. No personal data is sent to our server in any way. Nothing. Zilch. Nada. All the face detection algorithms will run on your Read more about How normal am I? – Let an AI judge you[…]
If you have ever dreamed of earning money from a stellar music career but were concerned you had little talent, don’t let that put you off – a man called Alex Mitchell might be able to help. Mr Mitchell is the founder and boss of a website and app called Boomy, which helps its users Read more about The AI software that could turn you in to a music star[…]
[…]DARPA’s Guaranteeing AI Robustness against Deception (GARD) program […] focuses on a few core objectives. One of which is the development of a testbed for characterizing ML defenses and assessing the scope of their applicability […] Ensuring that emerging defenses are keeping pace with – or surpassing – the capabilities of known attacks is critical Read more about DARPA Open Sources Resources to Aid Evaluation of Adversarial AI Defenses[…]
minDALL-E on Conceptual Captions minDALL-E, named after minGPT, is a 1.3B text-to-image generation model trained on 14 million image-text pairs for non-commercial purposes. Environment Setup Basic setup Other packages Model Checkpoint Model structure (two-stage autoregressive model) Stage1: Unlike the original DALL-E [1], we replace Discrete VAE with VQGAN [2] to generate high-quality samples effectively. We Read more about minDALL-E creates images based on text input[…]
If the Pentagon is going to rely on algorithms and artificial intelligence, it’s got to solve the problem of “brittle AI.” A top Air Force official recently illustrated just how far there is to go. In a recent test, an experimental target recognition program performed well when all of the conditions were perfect, but a Read more about This Air Force Targeting AI Thought It Had a 90% Success Rate. It Was More Like 25%[…]
[…] One of the first, and reportedly most widely used, is PredPol, its name an amalgamation of the words “predictive policing.” The software was derived from an algorithm used to predict earthquake aftershocks that was developed by professors at UCLA and released in 2011. By sending officers to patrol these algorithmically predicted hot spots, these Read more about How We Determined Predictive Policing Software Disproportionately Targeted Low-Income, Black, and Latino Neighborhoods[…]
A repository that shares tuning results of trained models generated by Tensorflow. Post-training quantization (Weight Quantization, Integer Quantization, Full Integer Quantization, Float16 Quantization), Quantization-aware training. I also try to convert it to OpenVINO’s IR model as much as possible. TensorFlow Lite, OpenVINO, CoreML, TensorFlow.js, TF-TRT, MediaPipe, ONNX [.tflite, .h5, .pb, saved_model, tfjs, tftrt, mlmodel, .xml/.bin, Read more about Tensorflow model zoo[…]
Intel today open-sourced ControlFlag, a tool that uses machine learning to detect problems in computer code — ideally to reduce the time required to debug apps and software. In tests, the company’s machine programming research team says that ControlFlag has found hundreds of defects in proprietary, “production-quality” software, demonstrating its usefulness. […] ControlFlag, which works Read more about Intel open-sources AI-powered tool to spot bugs in code[…]
Authorities in the United Arab Emirates have requested the US Department of Justice’s help in probing a case involving a bank manager who was swindled into transferring $35m to criminals by someone using a fake AI-generated voice. The employee received a call to move the company-owned funds by someone purporting to be a director from Read more about Criminals use fake AI voice to swindle UAE bank out of $35m[…]
Load up the website This Person Does Not Exist and it’ll show you a human face, near-perfect in its realism yet totally fake. Refresh and the neural network behind the site will generate another, and another, and another. The endless sequence of AI-crafted faces is produced by a generative adversarial network (GAN) — a type Read more about AI Fake-Face Generators Can Be Rewound To Reveal the Real Faces They Trained On[…]
[…] Humans should have full decision-making power, the guidelines state, and have the right to choose whether to accept AI services, exit an interaction with an AI system or discontinue its operation at any time. The document was published by China’s Ministry of Science and Technology (MOST) last Sunday. The goal is to “make sure Read more about Chinese AI gets ethical guidelines for the first time[…]
Last month, Stanford researchers declared that a new era of artificial intelligence had arrived, one built atop colossal neural networks and oceans of data. They said a new research center at Stanford would build—and study—these “foundation models” of AI. Critics of the idea surfaced quickly—including at the workshop organized to mark the launch of the Read more about A Stanford Proposal Over AI’s ‘Foundations’ Ignites Debate[…]
“OpenAI is the company running the text completion engine that makes you possible,” Jason Rohrer, an indie games developer, typed out in a message to Samantha. She was a chatbot he built using OpenAI’s GPT-3 technology. Her software had grown to be used by thousands of people, including one man who used the program to Read more about A developer used GPT-3 to build realistic custom personality AI chatbots. OpenAI shut it down. Wants content filters, privacy invasions and inability to model personalities.[…]
Boffins from Duke University say they have figured out a way to help protect artificial intelligences from adversarial image-modification attacks: by throwing a few imaginary numbers their way. […] The problem with reliability: adversarial attacks which modify the input imagery in a way imperceptible to the human eye. In an example from a 2015 paper Read more about Imaginary numbers help AIs solve the very real problem of adversarial imagery • The Register[…]
Neural networks can correctly guess a person’s race just by looking at their bodily x-rays and researchers have no idea how it can tell. There are biological features that can give clues to a person’s ethnicity, like the colour of their eyes or skin. But beneath all that, it’s difficult for humans to tell. That’s Read more about AI algorithms uncannily good at spotting your race from medical scans[…]
[…] The AI community, in particular, rushed to develop software that many believed would allow hospitals to diagnose or triage patients faster, bringing much-needed support to the front lines—in theory. In the end, many hundreds of predictive tools were developed. None of them made a real difference, and some were potentially harmful. That’s the damning Read more about Hundreds of AI tools have been built to catch covid. None of them helped.[…]
[…] The issue is that ambiguity or discrepancies can be introduced if the machine-learning software ignores certain invisible Unicode characters. What’s seen on screen or printed out, for instance, won’t match up with what the neural network saw and made a decision on. It may be possible abuse this lack of Unicode awareness for nefarious Read more about Hey, AI software developers, you are taking Unicode into account, right … right?[…]
In what can only be considered a triumph for all robot-kind, this week, a federal court has ruled that an artificially intelligent machine can, in fact, be an inventor—a decision that came after a year’s worth of legal battles across the globe. The ruling came on the heels of a years-long quest by University of Read more about Australian Court Rules That AI Can Be an Inventor, as does South Africa[…]
On May 31 last year, 25-year-old Safarain Herring was shot in the head and dropped off at St. Bernard Hospital in Chicago by a man named Michael Williams. He died two days later. Chicago police eventually arrested the 64-year-old Williams and charged him with murder (Williams maintains that Herring was hit in a drive-by shooting). Read more about Police Are Telling ShotSpotter to Alter Evidence From Gunshot-Detecting AI[…]
TikTok’s AI is no longer a secret — in fact, it’s now on the open market. The Financial Times has learned that parent company ByteDance quietly launched a BytePlus division that sells TikTok technology, including the recommendation algorithm. Customers can also buy computer vision tech, real-time effects and automated translations, among other features. BytePlus debuted Read more about TikTok’s AI is now available to other companies[…]