Transparency by Design: Closing the Gap Between ...
Transparency by Design: Closing the Gap Between Performance andInterpretability in Visual ReasoningDavid Mascharka 1Philip Tran2Ryan Soklaski1Arjun Majumdar 11MIT Lincoln Laboratory 2Planck Aerosystems question answering requires high-order reason-ing about an image, which is a fundamental capabilityneeded by machine systems to follow complex , modular networks have been shown to be an ef-fective framework for performing visual reasoning modular networks were initially designed with a de-gree of model Transparency , their performance on complexvisual reasoning benchmarks was lacking. Current state-of-the-art approaches do not provide an effective mecha-nism for understanding the reasoning process. In this paper,we close the performance gap Between interpretable modelsand state-of-the-art visual reasoning methods. We proposea set of visual-reasoning primitives which, when composed,manifest as a model capable of performing complex reason-ing tasks in an explicitly-interpretable manner.
Transparency by Design: Closing the Gap Between Performance and Interpretability in Visual Reasoning David Mascharka 1Philip Tran2 Ryan Soklaski Arjun Majumdar 1MIT Lincoln Laboratoryy 2Planck Aerosystemsz ffirst.lastg@ll.mit.edu, phil@planckaero.com
Download Transparency by Design: Closing the Gap Between ...
Information
Domain:
Source:
Link to this page:
Please notify us if you found a problem with this document:
Related search queries
Closing the Gap Between, Closing the Gap Between Strategy and Execution, CLOSING THE GAP, Amazon S3, Closing the Gap The gap between, Closing the achievement gap between, Closing the Gap Between ASIC and Custom: An ASIC, Closing the Gap Between ASIC and Custom: An ASIC Perspective, Subspace Pursuit for Compressive Sensing: Closing the Gap, Subspace Pursuit for Compressive Sensing: Closing the Gap Between, Between, Closing the theory-practice gap, The gap, NPQSL: Closing the Gap Module