Fig. 1

AI in materials characterization techniques and data analysis[8-13]"

Fig. 2

General processes of machine learning for the field of materials characterization[23-25]"

Fig. 3

The difference between traditional machine learning and deep learning"

Table 1

Machine learning models commonly used in the battery and materials domain"

Table 2

Conventional spectroscopic characterization methods and application examples"

Fig. 4

Application of AI in spectroscopic characterization techniques[33,40,96-97]"

Fig. 5

Active learning workflow to fit an MTP for both bulk and grain boundary structures of LSTZ0.75[99]"

Fig. 6

(a) Analysis of massive spectra data flow chart[40]; (b) Schematic diagram of benchmarking feature space for supervised machine learning[100]"

Fig. 7

(a) Ideal and actual EIS measurement[108]; (b) CNN for EIS reconstruction[104]; (c) Estimate DRT using EIS data[102]"

Fig. 8

(a) The schematic representation of information describing the experimental data, the descriptors and the machine learning methods involved in this study[97]; (b) Real-time high-precision measurement of LiPF6 and solvent (EC and EMC) concentrations in lithium-ion battery electrolyte using CNN[31]; (c) An AI-from-scratch (AI-ai) framework to predict SEI's XPS[32]"

Fig. 9

Artificial intelligence in imaging characterization[6,10-11,113]"

Fig. 10

Denoising and super-resolution of the Li maps by SC. High-quality (HQ) Li maps of the cathode film in the 0% charged, 100% charged, and 51% discharged states[114]"

Fig. 11

(a) The proposed machine-learning-assisted diffractive imaging data analysis[116]; (b) Schematic of 3D electrode microstructure generation using SliceGAN[120]"

Fig. 12

(a) Comparison of multiphase segmentation by different methods[122]; (b) Flowchart for quantitative understanding of the complex phases in layered cathodes enabled by HaU-Net phase segmentation[13]"