Matches in SemOpenAlex for { <https://semopenalex.org/work/W3195795911> ?p ?o ?g. }
Showing items 1 to 81 of
81
with 100 items per page.
- W3195795911 abstract "Deep neural networks are an attractive tool for compressing the control policy lookup tables in systems such as the Airborne Collision Avoidance System (ACAS). It is vital to ensure the safety of such neural controllers via verification techniques. The problem of analyzing ACAS Xu networks has motivated many successful neural network verifiers. These verifiers typically analyze the internal computation of neural networks to decide whether a property regarding the input/output holds. The intrinsic complexity of neural network computation renders such verifiers slow to run and vulnerable to floating-point error. This paper revisits the original problem of verifying ACAS Xu networks. The networks take low-dimensional sensory inputs with training data provided by a precomputed lookup table. We propose to prepend an input quantization layer to the network. Quantization allows efficient verification via input state enumeration, whose complexity is bounded by the size of the quantization space. Quantization is equivalent to nearest-neighbor interpolation at run time, which has been shown to provide acceptable accuracy for ACAS in simulation. Moreover, our technique can deliver exact verification results immune to floating-point error if we directly enumerate the network outputs on the target inference implementation or on an accurate simulation of the target implementation." @default.
- W3195795911 created "2021-08-30" @default.
- W3195795911 creator A5000699866 @default.
- W3195795911 creator A5045127387 @default.
- W3195795911 date "2021-08-18" @default.
- W3195795911 modified "2023-09-27" @default.
- W3195795911 title "Verifying Low-dimensional Input Neural Networks via Input Quantization" @default.
- W3195795911 cites W1964123482 @default.
- W3195795911 cites W2152494765 @default.
- W3195795911 cites W2545507732 @default.
- W3195795911 cites W2565186948 @default.
- W3195795911 cites W2594877703 @default.
- W3195795911 cites W2909765152 @default.
- W3195795911 cites W2963735478 @default.
- W3195795911 cites W2963755523 @default.
- W3195795911 cites W3009766965 @default.
- W3195795911 cites W3034726383 @default.
- W3195795911 cites W3043655933 @default.
- W3195795911 cites W3046272654 @default.
- W3195795911 cites W3105170486 @default.
- W3195795911 cites W3109541805 @default.
- W3195795911 cites W3137147200 @default.
- W3195795911 cites W587598985 @default.
- W3195795911 cites W3142308036 @default.
- W3195795911 hasPublicationYear "2021" @default.
- W3195795911 type Work @default.
- W3195795911 sameAs 3195795911 @default.
- W3195795911 citedByCount "1" @default.
- W3195795911 countsByYear W31957959112021 @default.
- W3195795911 crossrefType "posted-content" @default.
- W3195795911 hasAuthorship W3195795911A5000699866 @default.
- W3195795911 hasAuthorship W3195795911A5045127387 @default.
- W3195795911 hasConcept C11413529 @default.
- W3195795911 hasConcept C134306372 @default.
- W3195795911 hasConcept C134835016 @default.
- W3195795911 hasConcept C154945302 @default.
- W3195795911 hasConcept C199360897 @default.
- W3195795911 hasConcept C28855332 @default.
- W3195795911 hasConcept C33923547 @default.
- W3195795911 hasConcept C34388435 @default.
- W3195795911 hasConcept C41008148 @default.
- W3195795911 hasConcept C45374587 @default.
- W3195795911 hasConcept C50644808 @default.
- W3195795911 hasConceptScore W3195795911C11413529 @default.
- W3195795911 hasConceptScore W3195795911C134306372 @default.
- W3195795911 hasConceptScore W3195795911C134835016 @default.
- W3195795911 hasConceptScore W3195795911C154945302 @default.
- W3195795911 hasConceptScore W3195795911C199360897 @default.
- W3195795911 hasConceptScore W3195795911C28855332 @default.
- W3195795911 hasConceptScore W3195795911C33923547 @default.
- W3195795911 hasConceptScore W3195795911C34388435 @default.
- W3195795911 hasConceptScore W3195795911C41008148 @default.
- W3195795911 hasConceptScore W3195795911C45374587 @default.
- W3195795911 hasConceptScore W3195795911C50644808 @default.
- W3195795911 hasLocation W31957959111 @default.
- W3195795911 hasOpenAccess W3195795911 @default.
- W3195795911 hasPrimaryLocation W31957959111 @default.
- W3195795911 hasRelatedWork W1508860170 @default.
- W3195795911 hasRelatedWork W2029912145 @default.
- W3195795911 hasRelatedWork W2199532963 @default.
- W3195795911 hasRelatedWork W2340411233 @default.
- W3195795911 hasRelatedWork W2520372893 @default.
- W3195795911 hasRelatedWork W2555732556 @default.
- W3195795911 hasRelatedWork W2791367338 @default.
- W3195795911 hasRelatedWork W2807040120 @default.
- W3195795911 hasRelatedWork W2904181757 @default.
- W3195795911 hasRelatedWork W2912816250 @default.
- W3195795911 hasRelatedWork W2925104992 @default.
- W3195795911 hasRelatedWork W2971143061 @default.
- W3195795911 hasRelatedWork W3006292807 @default.
- W3195795911 hasRelatedWork W3016494263 @default.
- W3195795911 hasRelatedWork W3082707543 @default.
- W3195795911 hasRelatedWork W3131973483 @default.
- W3195795911 hasRelatedWork W3185335265 @default.
- W3195795911 hasRelatedWork W3196843416 @default.
- W3195795911 hasRelatedWork W3206076827 @default.
- W3195795911 hasRelatedWork W3181191750 @default.
- W3195795911 isParatext "false" @default.
- W3195795911 isRetracted "false" @default.
- W3195795911 magId "3195795911" @default.
- W3195795911 workType "article" @default.