Visualizing and interpreting representations learned by machine learning / deep learning algorithms is pretty interesting! As the saying goes -- "A picture is worth a thousand words", the same holds true with visualizations. A lot can be interpreted using the correct tools for visualization. In this post, I will cover some details on visualizing intermediate (hidden) layer features using dimension reduction techniques. We will work with the IMDB sentiment classification task (25000 training and 25000 test examples).

Join us for a live stream to your campus, with new content, key learnings, live Q&A and giveaways! Learn the fundamentals of Machine Learning, Deep Learning & Artificial Intelligence focused on newly optimized frameworks for Intel Architecture like Intel Software Optimization for neon, Intel Optimization for Caffe*, and Intel Optimization for TensorFlow* from our technical experts.

Artificial Intelligence is radically changing the way we think of technology. It is progressing rapidly, with key advancements ranging from virtual assistants (such as Apple's Siri and Microsoft Cortana) to fraud detection. This emerging tech now plays a part in everyday life. Another study performed by Forrester Research predicted an increase of 300% in investment in AI this year (2017), compared to last year. Tools and technologies play an important role in growth of any technology.

There is more to building a sustainable Deep Learning solution that what is provided by Deep Learning frameworks like TensorFlow and PyTorch. These frameworks are good enough for research, but they don't take into account the problems that crop up with production deployment. I've written previously about technical debt and the need from more adaptive biological like architectures. To support a viable business using Deep Learning, you absolutely need an architecture that supports sustainable improvement in the presence of frequent and unexpected changes in the environment. Current Deep Learning framework only provide a single part of a complete solution.

One of the major factors that will have a positive impact on the growth of this market includes the rising usage of deep learning technology among various industries such as automotive, advertisement, medical and others. Moreover, increasing acceptance of cloud based technology, high usage of deep learning in big data analytics, high R&D expansions for enhanced processing hardware for deep learning and rising applicability in healthcare and autonomous vehicles are fueling the market growth. However, rising difficulty in hardware owing to composite algorithm used in deep learning technology is acting as key barrier for the market. Moreover, the market has tremendous growth opportunity such as utilization of deep learning technology in smartphones and medical image analysis. On the other hand, setback in approval of neuromorphic technology for deep learning and development of algorithms' at a faster pace when compared with its hardware.

Deep convolutional neural networks (CNNs) based approaches are the state-of-the-art in various computer vision tasks, including face recognition. Considerable research effort is currently being directed towards further improving deep CNNs by focusing on more powerful model architectures and better learning techniques. However, studies systematically exploring the strengths and weaknesses of existing deep models for face recognition are still relatively scarce in the literature. In this paper, we try to fill this gap and study the effects of different covariates on the verification performance of four recent deep CNN models using the Labeled Faces in the Wild (LFW) dataset. Specifically, we investigate the influence of covariates related to: image quality -- blur, JPEG compression, occlusion, noise, image brightness, contrast, missing pixels; and model characteristics -- CNN architecture, color information, descriptor computation; and analyze their impact on the face verification performance of AlexNet, VGG-Face, GoogLeNet, and SqueezeNet. Based on comprehensive and rigorous experimentation, we identify the strengths and weaknesses of the deep learning models, and present key areas for potential future research. Our results indicate that high levels of noise, blur, missing pixels, and brightness have a detrimental effect on the verification performance of all models, whereas the impact of contrast changes and compression artifacts is limited. It has been found that the descriptor computation strategy and color information does not have a significant influence on performance.

Many applications in speech, robotics, finance, and biology deal with sequential data, where ordering matters and recurrent structures are common. However, this structure cannot be easily captured by standard kernel functions. To model such structure, we propose expressive closed-form kernel functions for Gaussian processes. The resulting model, GP-LSTM, fully encapsulates the inductive biases of long short-term memory (LSTM) recurrent networks, while retaining the nonparametric probabilistic advantages of Gaussian processes. We learn the properties of the proposed kernels by optimizing the Gaussian process marginal likelihood using a new provably convergent semi-stochastic gradient procedure, and exploit the structure of these kernels for scalable training and prediction. This approach provides a practical representation for Bayesian LSTMs. We demonstrate state-of-the-art performance on several benchmarks, and thoroughly investigate a consequential autonomous driving application, where the predictive uncertainties provided by GP-LSTM are uniquely valuable.

Deep learning can enable Internet of Things (IoT) devices to interpret unstructured multimedia data and intelligently react to both user and environmental events but has demanding performance and power requirements. The authors explore two ways to successfully integrate deep learning with low-power IoT products.

The Google AI Residency Program -- previously known as the Google Brain Residency Program -- is a 12-month research training role designed to jumpstart or advance your career in machine learning research. Residents will have the opportunity to work alongside distinguished scientists and engineers from various research teams. The Brain Residency Program was created in 2015 with the goal of training and supporting the next generation of deep learning researchers. With deep learning and other machine learning subfields fast becoming a critical area for a broad range of applications, people from a wide range of disciplines are beginning to realize the importance and impact of this area of research. With growing interest in the field, there is a corresponding need for researchers with hands-on experience in machine learning techniques and methodologies.

We first introduce the concept of driver DNA: a unique driving behavior profile, a fingerprint that is specific only to that driver. In order to derive that driver DNA we first need to extract meaningful features from individual car trips. Later on we aggregate features over many trips to derive a general driving profile. The source of the data is the mobile phone's accelerometer and gyroscope sensors. From the raw signal we derive the longitudinal and lateral accelerations of the car as well as its yaw or angular velocity in the vertical axis (Figure 1).