Category Archives: Abstract

Moving deep learning to the mass market

Category : Abstract

Speaker: Samer Hijazi, Ph.D., Engineering Director, Cadence IP group

Deep learning is becoming most widely used techniques for computer vision and pattern recognition. This primarily driven by it is outstanding recognition rate. However, it’s high computational requirements, drives the power requirements to levels that are not reasonable for most embedded systems.  In this talk, I will address this problem from two aspects. First, from problem definition and algorithm optimization point of view.  Second form target deployment platform selection and optimization prospective.

Cognitive Systems Architecture and Methodology

Category : Abstract

Speaker: Jeffrey T. Kreulen, Ph.D. IBM Watson Group, Distinguished Engineer

In this session, Jeff Kreulen will discuss a theoretical framework that uses human cognition to explain a continuum of adaptive and intelligent computational systems that enhance information organization, learning and understanding for the benefit and augmentation of people, business and the world.

Understanding human language, like a human. And other tales of cognitive computing

Category : Abstract

Speaker: Prof. Noah Goodman, Assistant Professor of Psychology, Computer Science (by courtesy), and Linguistics (by courtesy), Stanford University

Human cognition is incredibly flexible, partly because common-sense knowledge is uncertain but highly structured. Probabilistic programming languages (PPLs) provide a formal tool encompassing probabilistic uncertainty and compositional structure. I will show that PPLs allow us to model human language understanding as social reasoning grounded in structured common-sense knowledge. This framework captures vague adjectives (“Bob is tall’’), generic language (“boys are tall’’), hyperbole (“Bob is a hundred feet tall’’), and metaphor (“Bob is a giraffe’’). I will close with some implications this work has for the future of cognitive computing.

Rethinking computation: a processor architecture for artificial intelligence

Category : Abstract

Speaker: Amir Khosrowshahi, co-founder and CTO, Nervana Systems

Computers are increasingly being used to process and understand data. New approaches to computation are required as current processor technology reaches fundamental limits. Deep learning is a branch of machine learning that has recently achieved state-of-the-art in a wide range of domains including images, speech, and text. Nervana is developing a novel processor architecture for deep learning. By integrating the necessary distributed computational primitives at a low level into a new processor design, we are able to outperform current technology such as CPUs and GPUs by a large margin in speed, scaling, and efficiency.

Sensing and Sensibility— a Quest to Visual Intelligence

Category : Abstract

Speaker: Silvio Savarese, Assistant Professor, Computer Science Department, Stanford University

We are moving into a world where sensors will be everywhere and will be ubiquitous. In the recent years, we have seen an explosion of new artificial visual sensors that can integrate luminance values with other sensing modalities: infrared, thermal, gravity, depth, to cite a few. Sensing is not the hard problem here, however. Sensibility, or, intelligent understanding of the sensing data is the challenge. When we look at an environment such as a coffee shop, we don’t just recognize the objects in isolation, but rather perceive a rich scenery of the 3D space, its objects, the people and all the relations among them. This allows us to effortlessly navigate through the environment, or to interact with objects in the scene with amazing precision or to predict what is about to happen next. In this talk I will give an overview of the research from my group and discuss our latest work on designing visual models that can process different sensing modalities and enable sensibility. I will be also demonstrating that our models are potentially transformative in application areas related to autonomous or assisted navigation, smart environments, social robotics,  augmented reality, and large scale information management.

Understanding Machines

Category : Abstract

Speaker: Monica Anderson, CTO and co-founder of Sensai Corporation

The recent breakthroughs in Deep Learning enable human level precision in Voice Understanding systems in our cellphones and computers. This is just the first step in a progression towards general “Understanding Machines” which is replacing the traditional but impossible strategies to what we used to call “Artificial Intelligence”. I argue that personal Understanding Machines based on voice input and output – which I call “Confidantes” – are the most likely path to widespread adoption of such machines and that their adoption will improve our society and the human condition in major ways and provide a path that avoid the unemployment dystopias so frequently seen in the media.

Natural language conversations as an interface for smart machines

Category : Abstract

Speaker: Ilya Gelfenbeyn,CEO and co-founder of

Siri and Alexa and other assistants demonstrated that users are ready to talk to services and devices. Conversational user interfaces replace and augment GUI in different verticals, and in many cases serve as the only interface for devices. In this session Ilya will tell about main values that conversational interfaces provide for smart devices ecosystems and discuss challenges in designing conversational products.

Building Smart Natural Language Applications with Data Ninja Services

Category : Abstract

Speaker: Dr. Sayandev Mukherjee, Senior Research Engineer , DOCOMO Innovations Inc.

The Data Ninja suite of services was launched recently with the goal of enabling companies of any size to build advanced services with natural language processing, without having to hire a team of data scientists.

This talk will introduce the natural language text analytics and semantic analysis features of the Data Ninja suite of services, and illustrate their capabilities with an example of a News Explorer application.

Cognitive Computing: Augmenting Human Capability

Category : Abstract

Speaker: Dr. Jeffrey J. WelserVice President & Lab Director, IBM Research – Almaden

The purpose of technology has always been to help humans expand and scale their mental and physical capabilities. Up until recently, the primary purpose of computing technology focused on automating routine tasks and exponentially speeding up our capability to do precise computations on structured data. The increasing availability of unstructured, non-numerical data (social media, IoT, natural language documents, images, audio, video, etc.) however is creating huge new opportunities for building business value, but the challenges in dealing with the volume (and velocity, variety and veracity) of it is straining our current systems and architectures. A new era of “cognitive” systems is needed, which have the ability to parse through data and find patterns and connections more like the way our brains process this information — but at a massive scale. As the birthplace of Watson, IBM Research has been extending the boundaries of cognitive computing for over a decade. Since the launch of the original system, cognitive technologies have now been developed to assist professionals from oncologists diagnosing and treating cancer to marketers wanted to understand the underlying personality traits of their customers. We are working to create a foundational cognitive platform, broad and flexible enough to help companies transform their industries, and support applications from an ecosystem of developers. And beyond just advancing cognitive software technologies, we are also looking at new hardware architectures to expand cognitive capability out of the Cloud and into mobile devices and IoT sensors. In the emerging Cognitive Business era, all of these technologies will need to be combined with deep domain knowledge from data scientists, to insure the solutions are aimed at the right problems in each industry to achieve the highest impact.


Share on LinkedIn