Collaborating Authors


Deep Learning and Computer Vision A-Z : OpenCV, SSD & GANs


Deep Learning and Computer Vision A-Z: OpenCV, SSD & GANs, Become a Wizard of all the latest Computer Vision tools that exist out there. Detect anything and create powerful apps. You've definitely heard of AI and Deep Learning. But when you ask yourself, what is my position with respect to this new industrial revolution, that might lead you to another fundamental question: am I a consumer or a creator? For most people nowadays, the answer would be, a consumer.

How do we know AI is ready to be in the wild? Maybe a critic is needed


Mischief can happen when AI is let loose in the world, just like any technology. The examples of AI gone wrong are numerous, the most vivid in recent memory being the disastrously bad performance of Amazon's facial recognition technology, Rekognition, which had a propensity to erroneously match members of some ethnic groups with criminal mugshots to a disproportionate extent. Given the risk, how can society know if a technology has been adequately refined to a level where it is safe to deploy? "This is a really good question, and one we are actively working on," Sergey Levine, assistant professor with the University of California at Berkeley's department of electrical engineering and computer science, told ZDNet by email this week. Levine and colleagues have been working on an approach to machine learning where the decisions of a software program are subjected to a critique by another algorithm within the same program that acts adversarially.

Object Recognition by a Minimally Pre-Trained System in the Process of Environment Exploration


We update the method of describing and assessing the process of the study of an abstract environment by a system, proposed earlier. We do not model any biological cognition mechanisms and consider the system as an agent equipped with an information processor (or a group of such agents), which makes a move in the environment, consumes information supplied by the environment, and gives out the next move (hence, the process is considered as a game). The system moves in an unknown environment and should recognize new objects located in it. In this case, the system should build comprehensive images of visible things and memorize them if necessary (and it should also choose the current goal set). The main problems here are object recognition, and the informational reward rating in the game.

Guided-TTS: Text-to-Speech with Untranscribed Speech - Technology Org


Neural text-to-speech (TTS) models are successfully used to generate high-quality human-like speech. However, most TTS models can be trained if only the transcribed data of the desired speaker is given. That means that long-form untranscribed data, such as podcasts, cannot be used to train existing models. A recent paper on arXiv proposes an unconditional diffusion-based generative model. It is trained on untranscribed data that leverages a phoneme classifier for text-to-speech synthesis.

#AI ( #人工知能 )と #画像認識技術 の進化とニュースまとめ


かき揚げうどんと言えば 僕ってイメージが付いた イーンスパイアの横田です。 昨夜も群馬の高崎駅で食べた ワサビ入り野沢菜かき揚げ、 旨かったですよ(笑

Know the Edge AI Ecosystem


Successful adoption of Edge AI requires understanding and integrating different elements in a way that this stack can be seamlessly deployed in the target environment. Implementing an Edge AI application requires an understanding of aspects like the tasks to be performed, hardware, frameworks, and models. For deep neural networks to run at the edge; hardware, frameworks, and tools need to work collectively. As edge AI applications vary according to the use case, these requirements need to be thought through for each of the scenarios. It is necessary to select proper hardware, frameworks, and tools that will be compatible with each other and the best suited for the use case. Below we discuss briefly a few of the frameworks, hardware processors, and development boards.

Our global agreement on AI could reduce bias and surveillance

New Scientist

Artificial intelligence is more present in our lives than ever: it predicts what we want to say in emails, helps us navigate from A to B and improves our weather reports. The unprecedented speed with which vaccines for covid-19 were developed can also partly be attributed to the use of AI algorithms that rapidly crunched the data from numerous clinical trials, allowing researchers around the world to compare notes in real time. The data sets used to build AI often aren't representative of the diversity of the population, so it can produce discriminatory practices or biases. One example is facial recognition technology. This is used to access our mobile phones, bank accounts and apartment buildings, and is increasingly employed by police forces.

How computer vision technology is enabling micro-fulfilment - Logistics Business Magazine


Retailers are now adopting micro-fulfilment strategies for instant consumer gratification and improved product accessibility, as a competitive advantage. The supply chain industry grew during COVID-19 crisis and so did the need for faster operational processes and automation of human tasks. As part of it, the logistics sector is struggling to meet the growing consumer demands, high labour costs, regulatory measures, and siloed data, whilst complying with a dynamic environment. Complexities woven in the industry are not just occasional but tend to create a ripple effect across the infrastructure. Ultimately, the warehouse workforce strives to meet customers' requirements by managing incoming orders through multiple layers, regardless of inventory processes.

Attention Maps for Visualizing the Recaptured Image Classification Models


At Cermati and Indodana, we are developing a wide spectrum of fraud detection techniques for our financial technology (FinTech) application. The purpose is to thwart people with nefarious intent from defrauding the system. Many machine learning and image processing algorithms are used, including image segmentation, optical character recognition and texture classification, with use cases ranging from prefilling web forms to font type recognition. Deep learning is one of the advanced algorithms we use to classify fake and invalid documents. In this article, we will present how the attention map may be used to visualize the deep learning image classification model that we have trained.