SRI International
-
Evaluating Visual-Semantic Explanations using a Collaborative Image Guessing Game
Abstract While there have been many proposals on making AI algorithms explainable, few have attempted to evaluate the impact of AI-generated explanations on human performance in conducting human-AI collaborative tasks. To bridge the gap, we propose a Twenty-Questions style collaborative image retrieval game, Explanation-assisted Guess Which (ExAG), as a method of evaluating the efficacy of…
-
Next-Generation Spray Delivery of Vaccines and Other Therapeutics with High Viscosity Aerosol
While conventional methods for spraying (Spray 1.0) such as pump sprays are largely viscosity-limited and only work for fluids with consistency close to water, the PARC technology called FEA, short for Filament Extension Atomizer (Spray 2.0) can generate narrowly-dispersed micron-sized droplets from fluids with a wide range of viscosities, from water-like (1 mPa-s) to peanut…
-
How to train your speaker embedding extractor
In this study, we aim to explore some of the fundamental requirements for building a good speaker embeddings extractor. We analyze the impact of voice activity detection, types of degradation, the amount of degraded data, and number of speakers required for a good network.
-
Approaches to multi-domain language recognition
Approaches found to provide robustness in multi-domain LID include a domain-and-language-weighted Gaussian backend classifier, duration-aware calibration, and a source normalized multi-resolution neural network backend.
-
Voices Obscured in Complex Environmental Settings (VOiCES) corpus
This work is a multi-organizational effort led by SRI International and Lab41 with the intent to push forward state-of-the-art distant microphone approaches in signal processing and speech recognition.
-
Ethical Considerations for AI Researchers
While the ethics of AI is not clear-cut, there are guidelines we can consider to minimize the harm we might introduce.
-
Launching Drishti: Continuing SRI’s human-centric legacy with AI in the factory
By Prasad Akella, Founder and CEO, Drishti The next time you come to a meeting at SRI International’s headquarters, arrive early. Linger in front of the glass-enclosed exhibit that lines the lobby’s left wall. It contains descriptions and artifacts from SRI’s storied history, arranged thematically from the 40’s to today. You’ll learn about everything from radar…
-
Estimation of Graphlet Counts in Massive Networks
In this work, we propose an unbiased graphlet estimation framework that is (a) fast with significant speedups compared to the state-of- the-art, (b) parallel with nearly linear-speedups, (c) accurate with
-
III-Nitride Lasers Challenges and Approaches to Expand the Emission Wavelength into the UV
In this tutorial talk we will review III-Nitride-based laser technology, discuss materials challenges using high band gap semiconductors, and present different approaches and tricks to overcome these barriers (i.e., polarization doping, e-beam excitation and frequency multiplication).
-
The Evidence Based Curriculum Design Framework: Leveraging Diverse Perspectives in the Design Process
This paper describes our approach to integrating the learning sciences with best practices in app design: a design framework that involves researchers and developers in a co-development process to create apps based on research and evidence.
-
The impact of career and technical education on students with disabilities
We use multiple years of administrative data from Massachusetts to estimate the effect of participating in CTE on the academic outcomes of students with disabilities.
-
Chicago’s Pay for Success Program Evaluation: Year 3
For each year of the program, SRI International, the program’s independent evaluator, releases a report of student performance across kindergarten readiness, special education utilization, and third-grade literacy.