With all of the progress we've seen in deep learning tech in the past few years, it seems pretty inevitable that security cameras become smarter and more capable in regards to tracking, but there are more options than we think in how we choose to pull this off. Traces AI is a new computer vision startup, in Y Combinator's latest batch of bets, that's focused on helping cameras track people without relying on facial recognition data, something the founders believe is too invasive of the public's privacy. We can use your hair style, whether you have a backpack, your type of shoes and the combination of your clothing," co-founder Veronika Yurchuk tells TechCrunch. Tech like this obviously doesn't scale too well for a multi-day city-wide manhunt, and leaves room for some Jason Bourne-esque criminals to turn their jackets inside out and toss on a baseball cap to evade detection. As a potential customer, why forego a sophisticated technology just to stave off dystopia? Well, Traces AI isn't so convinced that facial recognition tech is always the best solution; they believe that facial tracking isn't something every customer wants or needs and there should be more variety in terms of solutions. "The biggest concern [detractors] have is, 'Okay, you want to ban the technology that is actually protecting people today, and will be protecting this country tomorrow?'

We need a lot of data thas a huge drawback. We are learning the representation directly and this is why it works so well. Even in RL, we need a lot of data and this can really be a drawback in every turn. And all supervised learning is backpropagation gradient descent derivative. Even 2005 some good progress were made.

In recent years, a growing number of researchers have explored the use of robotic arms or dexterous hands to solve a variety of everyday tasks. While many of them have successfully tackled simple tasks, such as grasping or basic manipulation, complex tasks that involve multiple steps and precise/strategic movements have so far proved harder to address. A team of researchers at the Chinese University of Hong Kong and Tencent AI Lab has recently developed a deep learning-based approach to solve a Rubik's Cube using a multi-fingered dexterous hand. Their approach, presented in a paper pre-published on arXiv, allows a dexterous hand to solve more advanced in-hand manipulation tasks, such as the renowned Rubik's Cube puzzle. A Rubik's Cube is a plastic cube covered in multi-colored squares that can be shifted into different positions.

Working with unstructured text data is hard especially when you are trying to build an intelligent system which interprets and understands free flowing natural language just like humans. You need to be able to process and transform noisy, unstructured textual data into some structured, vectorized formats which can be understood by any machine learning algorithm. Principles from Natural Language Processing, Machine Learning or Deep Learning all of which fall under the broad umbrella of Artificial Intelligence are effective tools of the trade. Based on my previous posts, an important point to remember here is that any machine learning algorithm is based on principles of statistics, math and optimization. Hence they are not intelligent enough to start processing text in their raw, native form.

After breaking all the records related to training computer vision models, NVIDIA now claims that it's AI platform is able to train a natural language neural network model based on one of the largest datasets in a record time. It also claims that the inference time is just 2 milliseconds which translates to an extremely fast response from the model participating in a conversation with a user. After computer vision, natural language processing is one of the top applications of AI. From Siri to Alexa to Cortana to Google Assistant, all conversational user experiences are powered by AI. The advancements in AI research is putting the power of language understanding and conversational interface into the hands of developers.

We start with a brief introduction to reinforcement learning (RL), about its successful stories, basics, an example, issues, the ICML 2019 Workshop on RL for Real Life, how to use it, study material and an outlook. Then we discuss a selection of RL applications, including recommender systems, computer systems, energy, finance, healthcare, robotics, and transportation.

Recent applications of machine learning algorithms in the seismic domain have shown great potential in different areas such as seismic inversion and interpretation. However, such algorithms rarely enforce geophysical constraints - the lack of which might lead to undesirable results. To overcome this issue, we have developed a semi-supervised sequence modeling framework based on recurrent neural networks for elastic impedance inversion from multi-angle seismic data. Specifically, seismic traces and elastic impedance (EI) traces are modeled as a time series. Then, a neural-network-based inversion model comprising convolutional and recurrent neural layers is used to invert seismic data for EI. The proposed workflow uses well-log data to guide the inversion. In addition, it uses seismic forward modeling to regularize the training and to serve as a geophysical constraint for the inversion. The proposed workflow achieves an average correlation of 98% between the estimated and target EI using 10 well logs for training on a synthetic data set.

Despite achieving remarkable success in various domains, recent studies have uncovered the vulnerability of deep neural networks to adversarial perturbations, creating concerns on model generalizability and new threats such as prediction-evasive misclassification or stealthy reprogramming. Among different defense proposals, stochastic network defenses such as random neuron activation pruning or random perturbation to layer inputs are shown to be promising for attack mitigation. However, one critical drawback of current defenses is that the robustness enhancement is at the cost of noticeable performance degradation on legitimate data, e.g., large drop in test accuracy. This paper is motivated by pursuing for a better trade-off between adversarial robustness and test accuracy for stochastic network defenses. We propose Defense Efficiency Score (DES), a comprehensive metric that measures the gain in unsuccessful attack attempts at the cost of drop in test accuracy of any defense. To achieve a better DES, we propose hierarchical random switching (HRS), which protects neural networks through a novel randomization scheme. A HRS-protected model contains several blocks of randomly switching channels to prevent adversaries from exploiting fixed model structures and parameters for their malicious purposes. Extensive experiments show that HRS is superior in defending against state-of-the-art white-box and adaptive adversarial misclassification attacks. We also demonstrate the effectiveness of HRS in defending adversarial reprogramming, which is the first defense against adversarial programs. Moreover, in most settings the average DES of HRS is at least 5X higher than current stochastic network defenses, validating its significantly improved robustness-accuracy trade-off.

The influence of deep learning is continuously expanding across different domains, and its new applications are ubiquitous. The question of neural network design thus increases in importance, as traditional empirical approaches are reaching their limits. Manual design of network architectures from scratch relies heavily on trial and error, while using existing pretrained models can introduce redundancies or vulnerabilities. Automated neural architecture design is able to overcome these problems, but the most successful algorithms operate on significantly constrained design spaces, assuming the target network to consist of identical repeating blocks. We propose a probabilistic representation of a neural network structure under the assumption of independence between layer types. The probability matrix (prototype) can describe general feedforward architectures and is equivalent to the population of models, while being simple to interpret and analyze. We construct an architecture search algorithm, inspired by the estimation of distribution algorithms, to take advantage of this representation. The probability matrix is tuned towards generating high-performance models by repeatedly sampling the architectures and evaluating the corresponding networks. Our algorithm is shown to discover models which are competitive with those produced by existing architecture search methods, both in accuracy and computational costs, despite the conceptual simplicity and the comparatively limited scope of achievable designs.

A critical part of multi-person multi-camera tracking is person re-identification (re-ID) algorithm, which recognizes and retains identities of all detected unknown people throughout the video stream. Many re-ID algorithms today exemplify state of the art results, but not much work has been done to explore the deployment of such algorithms for computation and power constrained real-time scenarios. In this paper, we study the effect of using a light-weight model, MobileNet-v2 for re-ID and investigate the impact of single (FP32) precision versus half (FP16) precision for training on the server and inference on the edge nodes. We further compare the results with the baseline model which uses ResNet-50 on state of the art benchmarks including CUHK03, Market-1501, and Duke-MTMC. The MobileNet-V2 mixed precision training method can improve both inference throughput on the edge node, and training time on server $3.25\times$ reaching to 27.77fps and $1.75\times$, respectively and decreases power consumption on the edge node by $1.45\times$, while it deteriorates accuracy only 5.6\% in respect to ResNet-50 single precision on the average for three different datasets. The code and pre-trained networks are publicly available at https://github.com/TeCSAR-UNCC/person-reid.