As they work to capitalize on voluminous amounts of data generated in the course of day-to-day business, business and IT leaders are increasingly turning to applications driven by artificial intelligence. These forward-looking leaders recognize that AI enables organizations to make big decisions and bold moves based on data-driven insights.
With those thoughts in mind, let’s look at three recent Dell Technologies papers that highlight the amazing potential of AI-driven applications, along with some of the deep learning techniques and technologies that will get us there.
With deep learning technologies, organizations can streamline and accelerate image and video classification applications. These AI-driven applications use computer vision to classify or categorize an image or video file on the basis of its visual content.
In this white paper, we highlight diverse use cases for putting these powerful applications to work to carry out tasks that would be extremely burdensome, if not impossible, with manual processes. To make this story more tangible, the paper includes examples of the ways in which organizations are leveraging AI-driven image and video classification in the course of their operations — to get better at the things they do best.
A few examples: With the right applications in place, a bank might use image identification to recognize its customers as they walk through the door. A retailer might use image classification to enable a checkout-free store. And a social media company might use video classification to categorize and annotate content. The list of potential applications of image and video classification goes on and on. And one commonality that these applications share is a foundation based on deep learning, one of the key building blocks for AI solutions.
The art of the possible
Dell Technologies has an active research program focused on helping organizations explore, develop and adopt natural language processing applications. This research is carried out by a data sciences team in the Dell Technologies HPC & AI Innovation Lab in Austin, Texas. This white paper explores two of the groundbreaking projects under way in the lab. One focuses on language-to-language translation and the other focuses on text-to-voice translation.
In the lab’s research project focused on language-to-language translation, our data scientists are working to solve key problems associated with translating from one human language to another using a neural network. This is a process that involves taking inputs from a source language and converting it to a target language.
In this process, the translation model first reads a sentence in a source language and then passes it to an encoder, which builds an intermediate representation. This intermediate representation is then passed to a decoder, which processes the intermediate representation to produce the translated sentence in the target language.
Text-to-voice translation takes written words and converts them to audio. The objective is to generate a complete audio wave form synthetically — while not using the mechanized, clip recordings that we have been used to hearing on telephone systems for the last 20 years.
With these more advanced approaches, developers use training data that consists of a transcript and clips of a voice actor reading that transcript. These resources serve as the training foundation for the creation of a voice that a computer will mimic. The developers then train the neural network to produce a voice that sounds extremely similar to the actor’s voice, although it’s not that person speaking. It’s a neural network creating that voice completely from scratch.
At Dell Technologies, we expect the retail world and its use cases for recommendation engines to be dramatically transformed by advances in AI in the coming years. To further these advances, our HPC & AI Innovation Lab is working actively to make the process of training AI algorithms faster and more efficient on Dell EMC infrastructure in order to build better recommendations engines.
In tests conducted in the lab, our data scientists were able to show that organizations can parallelize the training of the type of neural networks used to make recommendation engines. They further showed that with the approaches the lab is pioneering, organizations can achieve very good recommendations very quickly, even with large datasets.
This isn’t theoretical research. The lab team is demonstrating the foundation for applications that can be developed and deployed today to help retailers, content providers and others build better engines for driving their customers to consumer products, music, video, books and countless other offerings.
To learn more
To explore additional use cases for AI in the enterprise, see the customer stories on the Dell Technologies Artificial Intelligence Solutions site.