Information Sciences Seminar——BDD4BNN: A BDD-Based Quantitative Analysis Framework for Binarized Neural Networks
报告人：Fu Song (ShanghaiTech University)
时间：2021-09-16 10:00-12:00
地点：腾讯会议（会议 ID：110 764 650）
Abstract: Verifying and explaining the behavior of neural networks is becoming increasingly important, especially when they are deployed in safety-critical applications. In this paper, we study verification and interpretability problems for Binarized Neural Networks (BNNs), the 1-bit quantization of general real-numbered neural networks. Our approach is to encode BNNs into Binary Decision Diagrams (BDDs), which is done by exploiting the internal structure of the BNNs. In particular, we translate the input-output relation of blocks in BNNs to cardinality constraints which are in turn encoded by BDDs. Based on the encoding, we develop a quantitative framework for BNNs where precise and comprehensive analysis of BNNs can be performed. We demonstrate the application of our framework by providing quantitative robustness analysis and interpretability for BNNs. We implement a prototype tool BDD4BNN and carry out extensive experiments, confirming the effectiveness and efficiency of our approach.
Bio: 宋富是上海科技大学常任副教授，研究员，博士生导师，系统与安全中心主任，主要研究系统与软件安全验证和测试技术、及相关逻辑和自动机理论。宋富于2013年获巴黎狄德罗大学博士学位，曾在华东师范大学担任讲师和副研究员。主持和参与多项国家自然科学基金委青年、面上和重点项目，曾获上海市浦江人才和上海市晨光学者人才计划资助，已在国际著名会议或期刊（如CAV、TACAS、FM、AAAI、IJCAI、I&C）发表多篇论文。
腾讯会议
会议 ID：110 764 650
链接：https://meeting.tencent.com/dm/hRgugTHrrHiV
