An official website of the United States government
Colloidal self-assembly is a viable solution to making advanced metamaterials. While the physicochemical properties of the particles affect the properties of the assembled structures, particle configuration is also a critical determinant factor. Colloidal self-assembly state classification is typically achieved with order parameters, which are aggregate variables normally defined with nontrivial exploration and validation. Here, we present an image-based framework to classify the state of a 2-D colloidal self-assembly system. The framework leverages deep learning algorithms with unsupervised learning for state classification and a supervised learning-based convolutional neural network for state prediction. The neural network models are developed using data from an experimentally validated Brownian dynamics simulation. Our results demonstrate that the proposed approach gives a satisfying performance, comparable and even outperforming the commonly used order parameters in distinguishing void defective states from ordered states. Given the data-based nature of the approach, we anticipate its general applicability and potential automatability to different and complex systems where image or particle coordination acquisition is feasible.
more »
« less
Siamese neural network improves the performance of a convolutional neural network in colloidal self-assembly state classification
Identifying the state of the colloidal self-assembly process is critical to monitoring and controlling the system into desired configurations. Recent application of convolutional neural networks with unsupervised clustering has shown a comparable performance to conventional approaches, in representing and classifying the states of a simulated 2D colloidal batch assembly system. Despite the early success, capturing the subtle differences among similar configurations still presents a challenge. To address this issue, we leverage a Siamese neural network to improve the accuracy of the state classification. Results from a Brownian dynamics-simulated electric field-mediated colloidal self-assembly system and a magnetic field-mediated colloidal self-assembly system demonstrate significant improvement from the original convolutional neural network-based approach. We anticipate the proposed improvement to further pave the way for automated monitoring and control of colloidal self-assembly processes in real time and real space.
more »
« less
- Award ID(s):
- 2218077
- PAR ID:
- 10617449
- Publisher / Repository:
- Journal of Chemical Physics
- Date Published:
- Journal Name:
- The Journal of Chemical Physics
- Volume:
- 161
- Issue:
- 20
- ISSN:
- 0021-9606
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Gagliardi, Laura (Ed.)Colloidal particles with anisotropic geometries and interactions display rich phase behavior and hence have the potential to serve as the basis of functional materials, which can tunably and reversibly self-assemble into different configurations. External fields are one design parameter that can be used to manipulate how systems of colloidal particles assemble with one another. One challenge in designing new materials using anisotropic colloidal particles is understanding how an individual particle’s various anisotropic features, like geometry, affect their overall self-assembly. Here, we present the results of simulation studies that explore the self-assembly of 2D colloidal squares with offset magnetic dipoles in the presence of an external field. Annealing simulations are used to measure the equilibrium-phase behavior of systems of these particles in the ground state, when the magnetic interactions dominate over the thermal forces of the system. We find that the magnetic properties of these systems are strongly influenced by the relative number of squares with opposite “handedness”, or chirality, that are present within the system. Systems of squares that contain equal numbers of either chirality are extremely responsive to the external field; a relatively weak external field is required to magnetize them. In contrast, systems that contain only one chirality of squares are significantly less responsive to the external field; a significantly stronger external field is required to elicit the same magnetic response. Ultimately, the differing macroscopic magnetic properties of these systems are related to their microscopic self- assembly in an external field. Simulation snapshots and ground state phase diagrams illustrate how the absence of opposite chirality squares prevents systems of these particles from leaving an energetically favorable antiparallel configuration in the presence of an external field. When opposite chirality squares are present, these magnetic particles assemble into a head-to-tail configuration, therefore inducing a magnetic statemore » « less
-
Abstract Machine learning influences numerous aspects of modern society, empowers new technologies, from Alphago to ChatGPT, and increasingly materializes in consumer products such as smartphones and self-driving cars. Despite the vital role and broad applications of artificial neural networks, we lack systematic approaches, such as network science, to understand their underlying mechanism. The difficulty is rooted in many possible model configurations, each with different hyper-parameters and weighted architectures determined by noisy data. We bridge the gap by developing a mathematical framework that maps the neural network’s performance to the network characters of the line graph governed by the edge dynamics of stochastic gradient descent differential equations. This framework enables us to derive a neural capacitance metric to universally capture a model’s generalization capability on a downstream task and predict model performance using only early training results. The numerical results on 17 pre-trained ImageNet models across five benchmark datasets and one NAS benchmark indicate that our neural capacitance metric is a powerful indicator for model selection based only on early training results and is more efficient than state-of-the-art methods.more » « less
-
Background: Quantification of metabolites from nuclear magnetic resonance (NMR) spectra in an accurate, high-throughput manner requires effective data processing tools. Neural networks are relatively underexplored in quantitative NMR metabolomics despite impressive speed and throughput compared to more conventional peak-fitting metabolomics software. Methods: This work investigates practices for dataset and model development in the task of metabolite quantification directly from simulated NMR spectra for three neural network models: the multi-layered perceptron, the convolutional neural network, and the transformer. Model architectures, training parameters, and training datasets are optimized before comparing each model on simulated 400-MHz 1H-NMR spectra of complex mixtures with 8, 44, or 86 metabolites to quantify in spectra ranging from simple to highly complex and overlapping peaks. The optimized models were further validated on spectra at 100- and 800-MHz. Results: The transformer was the most effective network for NMR metabolite quantification, especially as the number of metabolites per spectra increased or target concentrations were low or had a large dynamic range. Further, the transformer was able to accurately quantify metabolites in simulated spectra from 100-MHz up to 800-MHz. Conclusions: The methods developed in this work reveal that transformers have the potential to accurately perform fully automated metabolite quantification in real-time and, with further development with experimental data, could be the basis for automated quantitative NMR metabolomics software.more » « less
-
We present DeepPicar, a low-cost deep neural network based autonomous car platform. DeepPicar is a small scale replication of a real self-driving car called DAVE-2 by NVIDIA. DAVE-2 uses a deep convolutional neural network (CNN), which takes images from a front-facing camera as input and produces car steering angles as output. DeepPicar uses the same network architecture-9 layers, 27 million connections and 250K parameters-and can drive itself in real-time using a web camera and a Raspberry Pi 3 quad-core platform. Using DeepPicar, we analyze the Pi 3's computing capabilities to support end-to-end deep learning based real-time control of autonomous vehicles. We also systematically compare other contemporary embedded computing platforms using the DeepPicar's CNN-based real-time control workload. We find that all tested platforms, including the Pi 3, are capable of supporting the CNN-based real-time control, from 20 Hz up to 100 Hz, depending on hardware platform. However, we find that shared resource contention remains an important issue that must be considered in applying CNN models on shared memory based embedded computing platforms; we observe up to 11.6X execution time increase in the CNN based control loop due to shared resource contention. To protect the CNN workload, we also evaluate state-of-the-art cache partitioning and memory bandwidth throttling techniques on the Pi 3. We find that cache partitioning is ineffective, while memory bandwidth throttling is an effective solution.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1063/5.0244337
