-
Refining Coded Image in Human Vision Layer Using CNN-Based Post-Processing
Authors:
Takahiro Shindo,
Yui Tatsumi,
Taiju Watanabe,
Hiroshi Watanabe
Abstract:
Scalable image coding for both humans and machines is a technique that has gained a lot of attention recently. This technology enables the hierarchical decoding of images for human vision and image recognition models. It is a highly effective method when images need to serve both purposes. However, no research has yet incorporated the post-processing commonly used in popular image compression sche…
▽ More
Scalable image coding for both humans and machines is a technique that has gained a lot of attention recently. This technology enables the hierarchical decoding of images for human vision and image recognition models. It is a highly effective method when images need to serve both purposes. However, no research has yet incorporated the post-processing commonly used in popular image compression schemes into scalable image coding method for humans and machines. In this paper, we propose a method to enhance the quality of decoded images for humans by integrating post-processing into scalable coding scheme. Experimental results show that the post-processing improves compression performance. Furthermore, the effectiveness of the proposed method is validated through comparisons with traditional methods.
△ Less
Submitted 16 June, 2024; v1 submitted 20 May, 2024;
originally announced May 2024.
-
Scalable Image Coding for Humans and Machines Using Feature Fusion Network
Authors:
Takahiro Shindo,
Taiju Watanabe,
Yui Tatsumi,
Hiroshi Watanabe
Abstract:
As image recognition models become more prevalent, scalable coding methods for machines and humans gain more importance. Applications of image recognition models include traffic monitoring and farm management. In these use cases, the scalable coding method proves effective because the tasks require occasional image checking by humans. Existing image compression methods for humans and machines meet…
▽ More
As image recognition models become more prevalent, scalable coding methods for machines and humans gain more importance. Applications of image recognition models include traffic monitoring and farm management. In these use cases, the scalable coding method proves effective because the tasks require occasional image checking by humans. Existing image compression methods for humans and machines meet these requirements to some extent. However, these compression methods are effective solely for specific image recognition models. We propose a learning-based scalable image coding method for humans and machines that is compatible with numerous image recognition models. We combine an image compression model for machines with a compression model, providing additional information to facilitate image decoding for humans. The features in these compression models are fused using a feature fusion network to achieve efficient image compression. Our method's additional information compression model is adjusted to reduce the number of parameters by enabling combinations of features of different sizes in the feature fusion network. Our approach confirms that the feature fusion network efficiently combines image compression models while reducing the number of parameters. Furthermore, we demonstrate the effectiveness of the proposed scalable coding method by evaluating the image compression performance in terms of decoded image quality and bitrate.
△ Less
Submitted 16 June, 2024; v1 submitted 15 May, 2024;
originally announced May 2024.
-
Single-Shot Laser-Driven Neutron Resonance Spectroscopy for Temperature Profiling
Authors:
Zechen Lan,
Yasunobu Arikawa,
S. Reza Mirfayzi,
Alessio Morace,
Takehito Hayakawa,
Hirotaka Sato,
Takashi Kamiyama,
Tianyun Wei,
Yuta Tatsumi,
Mitsuo Koizumi,
Yuki Abe,
Shinsuke Fujioka,
Kunioki Mima,
Ryosuke Kodama,
Akifumi Yogo
Abstract:
The temperature measurement of material inside of an object is one of the key technologies for control of dynamical processes. For this purpose, various techniques such as laser-based thermography and phase-contrast imaging thermography have been studied. However, it is, in principle, impossible to measure the temperature of an element inside of an object using these techniques. One of the possibl…
▽ More
The temperature measurement of material inside of an object is one of the key technologies for control of dynamical processes. For this purpose, various techniques such as laser-based thermography and phase-contrast imaging thermography have been studied. However, it is, in principle, impossible to measure the temperature of an element inside of an object using these techniques. One of the possible solutions is measurements of Doppler brooding effect in neutron resonance absorption (NRA). Here we present a method to measure the temperature of an element or an isotope inside of an object using NRA with a single neutron pulse of approximately 100 ns width provided from a high-power laser. We demonstrate temperature measurements of a tantalum (Ta) metallic foil heated from the room temperature up to 617 K. Although the neutron energy resolution is fluctuated from shot to shot, we obtain exactly the temperature using a reference of a silver (Ag) foil kept to the room temperature. A free gas model well reproduces the results. This method enables element(isotope)-sensitive thermometry to detect the instantaneous temperature rise in dynamical processes.
△ Less
Submitted 3 October, 2023; v1 submitted 2 October, 2023;
originally announced October 2023.
-
QERaman: An open-source program for calculating resonance Raman spectra based on Quantum ESPRESSO
Authors:
Nguyen T. Hung,
Jianqi Huang,
Yuki Tatsumi,
Teng Yang,
Riichiro Saito
Abstract:
We present an open-source program QERaman that computes first-order resonance Raman spectroscopy of materials using the output data from Quantum ESPRESSO. Complex values of Raman tensors are calculated based on the quantum description of the Raman scattering from calculations of electron-photon and electron-phonon matrix elements, which are obtained by using the modified Quantum ESPRESSO. Our prog…
▽ More
We present an open-source program QERaman that computes first-order resonance Raman spectroscopy of materials using the output data from Quantum ESPRESSO. Complex values of Raman tensors are calculated based on the quantum description of the Raman scattering from calculations of electron-photon and electron-phonon matrix elements, which are obtained by using the modified Quantum ESPRESSO. Our program also calculates the resonant Raman spectra as a function of incident laser energy for linearly- or circularly-polarized light. Hands-on tutorials for graphene and MoS$_2$ are given to show how to run QERaman. All codes, examples, and scripts are available on the GitHub repository.
△ Less
Submitted 10 August, 2023;
originally announced August 2023.
-
Plasma Concept for Generating Circularly Polarized Electromagnetic Waves with Relativistic Amplitude
Authors:
Takayoshi Sano,
Yusuke Tatsumi,
Masayasu Hata,
Yasuhiko Sentoku
Abstract:
Propagation features of circularly polarized (CP) electromagnetic waves in magnetized plasmas are determined by the plasma density and the magnetic field strength. This property can be applied to design a unique plasma photonic device for intense short-pulse lasers. We have demonstrated by numerical simulations that a thin plasma foil under an external magnetic field works as a polarizing plate to…
▽ More
Propagation features of circularly polarized (CP) electromagnetic waves in magnetized plasmas are determined by the plasma density and the magnetic field strength. This property can be applied to design a unique plasma photonic device for intense short-pulse lasers. We have demonstrated by numerical simulations that a thin plasma foil under an external magnetic field works as a polarizing plate to separate a linearly polarized laser into two CP waves traveling in the opposite direction. This plasma photonic device has an advantage for generating intense CP waves even with a relativistic amplitude. For various research purposes, intense CP lights are strongly required to create high energy density plasmas in the laboratory.
△ Less
Submitted 7 November, 2020;
originally announced November 2020.
