• Publications
  • Influence
History-Aware Data Structure Repair Using SAT
tl;dr
We present a novel technique, called history-aware contract-based repair for more efficient data structure repair using SAT. Expand
  • 15
  • 2
DeepSafe: A Data-driven Approach for Checking Adversarial Robustness in Neural Networks
tl;dr
We propose a novel approach for automatically identifying safe regions of the input space, within which the network is robust against adversarial perturbations. Expand
  • 42
  • 1
Symbolic Execution for Deep Neural Networks
tl;dr
We introduce DeepCheck, a new approach for validating DNNs based on core ideas from program analysis, specifically from symbolic execution. Expand
  • 31
  • 1
Data-guided repair of selection statements
tl;dr
We use semi-supervised learning to predict the correct behavior for defect-inducing data and by patching up any inaccuracies in the prediction by a SAT-based combinatorial search. Expand
  • 27
  • 1
Property Inference for Deep Neural Networks
  • D. Gopinath, Ankur Taly, Hayes Converse, Corina S. Pasareanu
  • Computer Science, Mathematics
  • 34th IEEE/ACM International Conference on…
  • 29 April 2019
tl;dr
We present techniques for automatically inferring formal properties of feed-forward neural networks. Expand
  • 9
  • 1
Plant Disease Detection using Image Processing
tl;dr
This paper aims to support and help the green house farmers in an efficient way.Identification of the plant diseases is the key to prevent the losses in the yield and quantity of the agricultural product. Expand
  • 6
  • 1
Finding Invariants in Deep Neural Networks
tl;dr
We present techniques to extract input invariants as convex predicates on the input space, and layer invariants that represent features captured in the hidden layers. Expand
  • 3
  • 1
Specification-Based Program Repair Using SAT
tl;dr
We present an automated approach for generating likely bug fixes using behavioral specifications using SAT to prune the ensuing nondeterminism and repair the faulty statement. Expand
  • 94
Improving the effectiveness of spectra-based fault localization using specifications
tl;dr
This paper presents a novel technique that applies spectra-based localization in synergy with specification-based analysis to more accurately locate faults. Expand
  • 28
DeepSafe: A Data-Driven Approach for Assessing Robustness of Neural Networks
tl;dr
We propose DeepSafe, a novel approach for automatically assessing the overall robustness of a neural network’s robustness against adversarial perturbations. Expand
  • 27