Abstract:The paper deals with the recognition of symmetric three-dimensional (3D) bodies that can be rotated and translated. We provide a complete list of all existing combinations of rotation and reflection symmetries in 3D. We define 3D complex moments by means of spherical harmonics, and the influence of individual symmetry groups on complex moment values is studied. Each particular symmetry pre-defines certain moment values. These moments can no longer differentiate between two objects of the same symmetry, which d… Show more
“…The current set of invariants are limited to a maximum order of n max = 7 (and therefore l max = 7), which effectively limits their discriminative power. This is known to be particularly problematic for highly symmetric environments which require a high angular frequency to be described correctly (in such environments, low order moments are often zero) [90]. The limitation of n max = 7 comes from numerical issues in correctly identifying independent invariants, as standard singular value decomposition methods operate on floating point numbers where a threshold for comparing numbers must be carefully chosen.…”
In this work we apply methods for describing three-dimensional images to the problem of encoding atomic environments in a way that is invariant to rotations, translations, and permutations of the atoms and, crucially, can be decoded back into the original environment modulo global orientation without the need for training a model. From the point of view of decoding, the descriptor is optimally complete and can be extended to arbitrary order, allowing for a systematic convergence of the fidelity of the description. In experiments on molecules ranging from 3 to 29 atoms in size, we demonstrate that positions can be decoded with a 97% success rate and positions plus species with a 70% rate of success, rising to 95% if a second fingerprint is used. In all cases, consistent recovery is observed for molecules with 17 or fewer atoms. Additionally, we evaluate the descriptor's performance in predicting the energies and forces of bulk Ni, Cu, Li, Mo, Si, and Ge by means of a neural network model trained on DFT data. When comparing to six machine learning interaction potential methods that use various descriptors and regression schemes, our descriptor is found to be competitive, in several cases outperforming well established methods. The combined ability to both decode and make property predictions from a representation that does not need to be learned lays the foundations for a novel way of building generative models that are tasked with solving the inverse problem of predicting atomic arrangements that are statistically likely to have certain desired properties.
“…The current set of invariants are limited to a maximum order of n max = 7 (and therefore l max = 7), which effectively limits their discriminative power. This is known to be particularly problematic for highly symmetric environments which require a high angular frequency to be described correctly (in such environments, low order moments are often zero) [90]. The limitation of n max = 7 comes from numerical issues in correctly identifying independent invariants, as standard singular value decomposition methods operate on floating point numbers where a threshold for comparing numbers must be carefully chosen.…”
In this work we apply methods for describing three-dimensional images to the problem of encoding atomic environments in a way that is invariant to rotations, translations, and permutations of the atoms and, crucially, can be decoded back into the original environment modulo global orientation without the need for training a model. From the point of view of decoding, the descriptor is optimally complete and can be extended to arbitrary order, allowing for a systematic convergence of the fidelity of the description. In experiments on molecules ranging from 3 to 29 atoms in size, we demonstrate that positions can be decoded with a 97% success rate and positions plus species with a 70% rate of success, rising to 95% if a second fingerprint is used. In all cases, consistent recovery is observed for molecules with 17 or fewer atoms. Additionally, we evaluate the descriptor's performance in predicting the energies and forces of bulk Ni, Cu, Li, Mo, Si, and Ge by means of a neural network model trained on DFT data. When comparing to six machine learning interaction potential methods that use various descriptors and regression schemes, our descriptor is found to be competitive, in several cases outperforming well established methods. The combined ability to both decode and make property predictions from a representation that does not need to be learned lays the foundations for a novel way of building generative models that are tasked with solving the inverse problem of predicting atomic arrangements that are statistically likely to have certain desired properties.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.