Set-to-set matching aims to identify correspondences between two sets of unordered items by minimizing a distance metric or maximizing a similarity measure. Traditional metrics, such as Chamfer Distance (CD) and Earth Mover’s Distance (EMD), are widely used for this purpose but often suffer from limitations like suboptimal performance in terms of accuracy and robustness, or high computational costs - or both. In this paper, we propose a novel, simple yet effective set-to-set matching similarity measure, GPS, based on Gumbel prior distributions. These distributions are typically used to model the extrema of samples drawn from various distributions. Our approach is motivated by the observation that the distributions of minimum distances from CD, as encountered in real world applications such as point cloud completion, can be accurately modeled using Gumbel distributions. We validate our method on tasks like few-shot image classification and 3D point cloud completion, demonstrating significant improvements over state of-the-art loss functions across several benchmark datasets. Our demo code is publicly available at https://github.com/Zhang-VISLab/ICLR2025-GPS

3D point clouds enhanced the robot’s ability to perceive the geometrical information of the environments, making it possible for many downstream tasks such as grasp pose detection and scene understanding. The performance of these tasks, though, heavily relies on the quality of data input, as incomplete can lead to poor results and failure cases. Recent training loss functions designed for deep learning-based point cloud completion, such as Chamfer distance (CD) and its variants (e.g. HyperCD [1]), imply a good gradient weighting scheme can significantly boost performance. However, these CD-based loss functions usually require data-related parameter tuning, which can be time-consuming for data-extensive tasks. To address this issue, we aim to find a family of weighted training losses (weighted CD) that requires no parameter tuning. To this end, we propose a search scheme, Loss Distillation via Gradient Matching, to find good candidate loss functions by mimicking the learning behavior in backpropagation between HyperCD and weighted CD. Once this is done, we propose a novel bilevel optimization formula to train the backbone network based on the weighted CD loss. We observe that: (1) with proper weighted functions, the weighted CD can always achieve similar performance to HyperCD, and (2) the Landau weighted CD, namely Landau CD, can outperform HyperCD for point cloud completion and lead to new state-of-the-art results on several benchmark datasets. Our demo code is available at https://github.com/Zhang-VISLab/IROS2024-LossDistillationWeightedCD.

Chamfer distance (CD) is a standard metric to measure the shape dissimilarity between point clouds in point cloud completion, as well as a loss function for (deep) learning. However, it is well known that CD is vulnerable to outliers, leading to the drift towards suboptimal models. In contrast to the literature where most works address such issues in Euclidean space, we propose an extremely simple yet powerful metric for point cloud completion, namely Hyperbolic Chamfer Distance (HyperCD), that computes CD in hyperbolic space. In backpropagation, HyperCD consistently assigns higher weights to the matched point pairs with smaller Euclidean distances. In this way, good point matches are likely to be preserved while bad matches can be updated gradually, leading to better completion results. We demonstrate state-of-the-art performance on the benchmark datasets, i.e. PCN, ShapeNet-55, and ShapeNet34, and show from visualization that HyperCD can significantly improve the surface smoothness. Code is available at: https://github.com/Zhang-VISLab.

A point cloud is a discrete set of data points sampled from a 3D geometric surface. Chamfer distance (CD) is a popular metric and training loss to measure the distances between point clouds, but also well known to be sensitive to outliers. To address this issue, in this paper we propose InfoCD, a novel contrastive Chamfer distance loss to learn to spread the matched points for better distribution alignments between point clouds as well as accounting for a surface similarity estimator. We show that minimizing InfoCD is equivalent to maximizing a lower bound of the mutual information between the underlying geometric surfaces represented by the point clouds, leading to a regularized CD metric which is robust and computationally efficient for deep learning. We conduct comprehensive experiments for point cloud completion using InfoCD and observe significant improvements consistently over all the popular baseline networks trained with CD-based losses, leading to new state-of-the-art results on several benchmark datasets. Demo code is available at https://github.com/Zhang-VISLab/NeurIPS2023-InfoCD.

Point clouds measured by 3D scanning devices often have partially missing data due to the view positioning of the scanner. The missing data can reduce the performance of a point cloud in downstream tasks such as segmentation, location, and pose estimation. Consequently, 3D point cloud completion aims to predict the missing regions of incomplete objects for these fundamental 3D vision tasks. However, predicting the complete object can easily diminish the detail or structure of a measured region, which usually does not require repair. This study proposes a novel neural network architecture, Cosmos Propagation Network (CP-Net), for 3D point cloud completion. CP-Net extracts latent features in different scales from incomplete point clouds used as input. For point cloud generation, we propose a novel point expand method using a Mirror Expand module. Compared with existing methods, our Mirror Expand module introduces less information redundancy, which makes the distribution of points more reliable. CP-Net predicts the details of missing regions and maintains a clear general structure. The performance of CP-Net on several benchmarks was compared to that of current baseline methods. Compared to the existing methods, CP-Net showed the best performance for various metrics. Thus, CP-Net is expected to help address various problems related to 3D point cloud completion. Its source code is available at https://github.com/ark1234/CP-Net.

Air pollution is a societal and cross-boundary environmental problem that can be visualized using a satellite. Satellite imaging is not only useful to the home country but also to the neighboring countries. Moreover, monitoring the movement of air pollution can help susceptible people avoid acid rain and photochemical smog. Using advanced remote sensing (RS) images, substantial information can be obtained, which can produce numerous effective methods for visualizing air pollution. In this article, a novel method for extracting air pollution has been proposed; it applies various pipeline networks along with a focus area method to exploit the spectral aspect information. Afterward, three indices with numerous modified fully convolutional networks (FCNs) were extracted. Then, by employing a multivote module, visualized air pollution can be presented. In the conducted experiments, five-year Himawari-8 satellite images have been utilized in the North–East Asia area to validate the frameworks. Furthermore, the experimental result indicating that the given methods could effectively visualize air pollution. Source code and data sets are available athttps://github.com/ark1234/Himawari-8-based-visualized-understanding.

This article describes several features in the MAFFT online service for multiple sequence alignment (MSA). As a result of recent advances in sequencing technologies, huge numbers of biological sequences are available and the need for MSAs with large numbers of sequences is increasing. To extract biologically relevant information from such data, sophistication of algorithms is necessary but not sufficient. Intuitive and interactive tools for experimental biologists to semiautomatically handle large data are becoming important. We are working on development of MAFFT toward these two directions. Here, we explain (i) the Web interface for recently developed options for large data and (ii) interactive usage to refine sequence data sets and MSAs.

Summary: We report an update for the MAFFT multiple sequence alignment program to enable parallel calculation of large numbers of sequences. The G-INS-1 option of MAFFT was recently reported to have higher accuracy than other methods for large data, but this method has been impractical for most large-scale analyses, due to the requirement of large computational resources. We introduce a scalable variant, G-large-INS-1, which has equivalent accuracy to G-INS-1 and is applicable to 50 000 or more sequences. Availability and implementation: This feature is available in MAFFT versions 7.355 or later at https://mafft.cbrc.jp/alignment/software/mpi.html. Supplementary information: Supplementary data are available at Bioinformatics online.

BACKGROUND: Long short-term memory (LSTM) is one of the most attractive deep learning methods to learn time series or contexts of input data. Increasing studies, including biological sequence analyses in bioinformatics, utilize this architecture. Amino acid sequence profiles are widely used for bioinformatics studies, such as sequence similarity searches, multiple alignments, and evolutionary analyses. Currently, many biological sequences are becoming available, and the rapidly increasing amount of sequence data emphasizes the importance of scalable generators of amino acid sequence profiles. RESULTS: We employed the LSTM network and developed a novel profile generator to construct profiles without any assumptions, except for input sequence context. Our method could generate better profiles than existing de novo profile generators, including CSBuild and RPS-BLAST, on the basis of profile-sequence similarity search performance with linear calculation costs against input sequence size. In addition, we analyzed the effects of the memory power of LSTM and found that LSTM had high potential power to detect long-range interactions between amino acids, as in the case of beta-strand formation, which has been a difficult problem in protein bioinformatics using sequence information. CONCLUSION: We demonstrated the importance of sequence context and the feasibility of LSTM on biological sequence analyses. Our results demonstrated the effectiveness of memories in LSTM and showed that our de novo profile generator, SPBuild, achieved higher performance than that of existing methods for profile prediction of beta-strands, where long-range interactions of amino acids are important and are known to be difficult for the existing window-based prediction methods. Our findings will be useful for the development of other prediction methods related to biological sequences by machine learning methods.

N-terminal acetylation is one of the most common protein modifications in eukaryotes and occurs co-translationally when the N-terminus of the nascent polypeptide is still attached to the ribosome. This modification has been shown to be involved in a wide range of biological phenomena such as protein half-life regulation, protein-protein and protein-membrane interactions, and protein subcellular localization. Thus, accurately predicting which proteins receive an acetyl group based on their protein sequence is expected to facilitate the functional study of this modification. As the occurrence of N-terminal acetylation strongly depends on the context of protein sequences, attempts to understand the sequence determinants of N-terminal acetylation were conducted initially by simply examining the N-terminal sequences of many acetylated and unacetylated proteins and more recently by machine learning approaches. However, a complete understanding of the sequence determinants of this modification remains to be elucidated. We obtained curated N-terminally acetylated and unacetylated sequences from the UniProt database and employed a decision tree algorithm to identify the sequence determinants of N-terminal acetylation for proteins whose initiator methionine (iMet) residues have been removed. The results suggested that the main determinants of N-terminal acetylation are contained within the first five residues following iMet and that the first and second positions are the most important discriminator for the occurrence of this phenomenon. The results also indicated the existence of position-specific preferred and inhibitory residues that determine the occurrence of N-terminal acetylation. The developed predictor software, termed NT-AcPredictor, accurately predicted the N-terminal acetylation, with an overall performance comparable or superior to those of preceding predictors incorporating machine learning algorithms. Our machine learning approach based on a decision tree algorithm successfully provided several sequence determinants of N-terminal acetylation for proteins lacking iMet, some of which have not previously been described. Although these sequence determinants remain insufficient to comprehensively predict the occurrence of this modification, indicating that further work on this topic is still required, the developed predictor, NT-AcPredictor, can be used to predict N-terminal acetylation with an accuracy of more than 80%.

MOTIVATION: Large multiple sequence alignments (MSAs), consisting of thousands of sequences, are becoming more and more common, due to advances in sequencing technologies. The MAFFT MSA program has several options for building large MSAs, but their performances have not been sufficiently assessed yet, because realistic benchmarking of large MSAs has been difficult. Recently, such assessments have been made possible through the HomFam and ContTest benchmark protein datasets. Along with the development of these datasets, an interesting theory was proposed: chained guide trees increase the accuracy of MSAs of structurally conserved regions. This theory challenges the basis of progressive alignment methods and needs to be examined by being compared with other known methods including computationally intensive ones. RESULTS: We used HomFam, ContTest and OXFam (an extended version of OXBench) to evaluate several methods enabled in MAFFT: (1) a progressive method with approximate guide trees, (2) a progressive method with chained guide trees, (3) a combination of an iterative refinement method and a progressive method and (4) a less approximate progressive method that uses a rigorous guide tree and consistency score. Other programs, Clustal Omega and UPP, available for large MSAs, were also included into the comparison. The effect of method 2 (chained guide trees) was positive in ContTest but negative in HomFam and OXFam. Methods 3 and 4 increased the benchmark scores more consistently than method 2 for the three datasets, suggesting that they are safer to use. AVAILABILITY AND IMPLEMENTATION: http://mafft.cbrc.jp/alignment/software/ CONTACT: [email protected] information: Supplementary data are available at Bioinformatics online.

MOTIVATION: Although many amino acid substitution matrices have been developed, it has not been well understood which is the best for similarity searches, especially for remote homology detection. Therefore, we collected information related to existing matrices, condensed it and derived a novel matrix that can detect more remote homology than ever. RESULTS: Using principal component analysis with existing matrices and benchmarks, we developed a novel matrix, which we designate as MIQS. The detection performance of MIQS is validated and compared with that of existing general purpose matrices using SSEARCH with optimized gap penalties for each matrix. Results show that MIQS is able to detect more remote homology than the existing matrices on an independent dataset. In addition, the performance of our developed matrix was superior to that of CS-BLAST, which was a novel similarity search method with no amino acid matrix. We also evaluated the alignment quality of matrices and methods, which revealed that MIQS shows higher alignment sensitivity than that with the existing matrix series and CS-BLAST. Fundamentally, these results are expected to constitute good proof of the availability and/or importance of amino acid matrices in sequence analysis. Moreover, with our developed matrix, sophisticated similarity search methods such as sequence-profile and profile-profile comparison methods can be improved further. AVAILABILITY AND IMPLEMENTATION: Newly developed matrices and datasets used for this study are available at http://csas.cbrc.jp/Ssearch/.