-
AI-optimized detector design for the future Electron-Ion Collider: the dual-radiator RICH case
Authors:
E. Cisbani,
A. Del Dotto,
C. Fanelli,
M. Williams,
M. Alfred,
F. Barbosa,
L. Barion,
V. Berdnikov,
W. Brooks,
T. Cao,
M. Contalbrigo,
S. Danagoulian,
A. Datta,
M. Demarteau,
A. Denisov,
M. Diefenthaler,
A. Durum,
D. Fields,
Y. Furletova,
C. Gleason,
M. Grosse-Perdekamp,
M. Hattawy,
X. He,
H. van Hecke,
D. Higinbotham
, et al. (22 additional authors not shown)
Abstract:
Advanced detector R&D requires performing computationally intensive and detailed simulations as part of the detector-design optimization process. We propose a general approach to this process based on Bayesian optimization and machine learning that encodes detector requirements. As a case study, we focus on the design of the dual-radiator Ring Imaging Cherenkov (dRICH) detector under development a…
▽ More
Advanced detector R&D requires performing computationally intensive and detailed simulations as part of the detector-design optimization process. We propose a general approach to this process based on Bayesian optimization and machine learning that encodes detector requirements. As a case study, we focus on the design of the dual-radiator Ring Imaging Cherenkov (dRICH) detector under development as part of the particle-identification system at the future Electron-Ion Collider (EIC). The EIC is a US-led frontier accelerator project for nuclear physics, which has been proposed to further explore the structure and interactions of nuclear matter at the scale of sea quarks and gluons. We show that the detector design obtained with our automated and highly parallelized framework outperforms the baseline dRICH design within the assumptions of the current model. Our approach can be applied to any detector R&D, provided that realistic simulations are available.
△ Less
Submitted 6 June, 2020; v1 submitted 13 November, 2019;
originally announced November 2019.
-
Mull it over: mutation testing based on LLVM
Authors:
Alex Denisov,
Stanislav Pankevich
Abstract:
This paper describes Mull, an open-source tool for mutation testing based on the LLVM framework. Mull works with LLVM IR, a low-level intermediate representation, to perform mutations, and uses LLVM JIT for just-in-time compilation. This design choice enables the following two capabilities of Mull: language independence and fine-grained control over compilation and execution of a tested program an…
▽ More
This paper describes Mull, an open-source tool for mutation testing based on the LLVM framework. Mull works with LLVM IR, a low-level intermediate representation, to perform mutations, and uses LLVM JIT for just-in-time compilation. This design choice enables the following two capabilities of Mull: language independence and fine-grained control over compilation and execution of a tested program and its mutations. Mull can work with code written in any programming language that supports compilation to LLVM IR, such as C, C++, Rust, or Swift. Direct manipulation of LLVM IR allows Mull to do less work to generate mutations: only modified fragments of IR code are recompiled, and this results in faster processing of mutated programs. To our knowledge, no existing mutation testing tool provides these capabilities for compiled programming languages. We describe the algorithm and implementation details of Mull, highlight current limitations of Mull, and present the results of our evaluation of Mull on real-world projects such as RODOS, OpenSSL, LLVM.
△ Less
Submitted 5 August, 2019;
originally announced August 2019.
-
On basis images for the digital image representation
Authors:
V. N. Gorbachev,
L. A. Denisov,
E. M. Kaynarova,
I. K. Metelev,
E. S. Yakovleva
Abstract:
Digital array orthogonal transformations that can be presented as a decomposition over basis items or basis images are considered. The orthogonal transform provides digital data scattering, a process of pixel energy redistributing, that is illustrated with the help of basis images. Data scattering plays important role for applications as image coding and watermarking. We established a simple quant…
▽ More
Digital array orthogonal transformations that can be presented as a decomposition over basis items or basis images are considered. The orthogonal transform provides digital data scattering, a process of pixel energy redistributing, that is illustrated with the help of basis images. Data scattering plays important role for applications as image coding and watermarking. We established a simple quantum analogues of basis images. They are representations of quantum operators that describe transition of single particle between its states.
Considering basis images as items of a matrix, we introduced a block matrix that is suitable for orthogonal transforms of multi-dimensional arrays such as block vector, components of which are matrices. We present an orthogonal transform that produces correlation between arrays. Due to correlation new feature of data scattering was found. A presented detection algorithm is an example of how it can be used in frequency domain watermarking.
△ Less
Submitted 23 January, 2019;
originally announced January 2019.
-
Fast Adjustable Threshold For Uniform Neural Network Quantization (Winning solution of LPIRC-II)
Authors:
Alexander Goncharenko,
Andrey Denisov,
Sergey Alyamkin,
Evgeny Terentev
Abstract:
Neural network quantization procedure is the necessary step for porting of neural networks to mobile devices. Quantization allows accelerating the inference, reducing memory consumption and model size. It can be performed without fine-tuning using calibration procedure (calculation of parameters necessary for quantization), or it is possible to train the network with quantization from scratch. Tra…
▽ More
Neural network quantization procedure is the necessary step for porting of neural networks to mobile devices. Quantization allows accelerating the inference, reducing memory consumption and model size. It can be performed without fine-tuning using calibration procedure (calculation of parameters necessary for quantization), or it is possible to train the network with quantization from scratch. Training with quantization from scratch on the labeled data is rather long and resource-consuming procedure. Quantization of network without fine-tuning leads to accuracy drop because of outliers which appear during the calibration. In this article we suggest to simplify the quantization procedure significantly by introducing the trained scale factors for quantization thresholds. It allows speeding up the process of quantization with fine-tuning up to 8 epochs as well as reducing the requirements to the set of train images. By our knowledge, the proposed method allowed us to get the first public available quantized version of MNAS without significant accuracy reduction - 74.8% vs 75.3% for original full-precision network. Model and code are ready for use and available at: https://github.com/agoncharenko1992/FAT-fast_adjustable_threshold.
△ Less
Submitted 18 June, 2019; v1 submitted 19 December, 2018;
originally announced December 2018.
-
On the steganographic image based approach to PDF files protection
Authors:
V. N. Gorbachev,
L. A. Denisov,
E. M. Kaynarova,
I. K. Metelev
Abstract:
Digital images can be copied without authorization and have to be protected. Two schemes for watermarking images in PDF document were considered. Both schemes include a converter to extract images from PDF pages and return the protected images back. Frequency and spatial domain embedding were used for hiding a message presented by a binary pattern. We considered visible and invisible watermarking…
▽ More
Digital images can be copied without authorization and have to be protected. Two schemes for watermarking images in PDF document were considered. Both schemes include a converter to extract images from PDF pages and return the protected images back. Frequency and spatial domain embedding were used for hiding a message presented by a binary pattern. We considered visible and invisible watermarking and found that spatial domain LSB technique can be more preferable than frequency embedding using DWT.
△ Less
Submitted 5 July, 2017;
originally announced July 2017.
-
Embedding of binary image in the Gray planes
Authors:
V. N. Gorbachev,
L. A. Denisov,
E. M. Kainarova
Abstract:
For watermarking of the digital grayscale image its Gray planes have been used. With the help of the introduced representation over Gray planes the LSB embedding method and detection have been discussed. It found that data, a binary image, hidden in the Gray planes is more robust to JPEG lossy compression than in the bit planes.
For watermarking of the digital grayscale image its Gray planes have been used. With the help of the introduced representation over Gray planes the LSB embedding method and detection have been discussed. It found that data, a binary image, hidden in the Gray planes is more robust to JPEG lossy compression than in the bit planes.
△ Less
Submitted 28 January, 2015;
originally announced January 2015.
