Alert button
Picture for Makoto Yamada

Makoto Yamada

Alert button

Implicit neural representation for change detection

Jul 28, 2023
Peter Naylor, Diego Di Carlo, Arianna Traviglia, Makoto Yamada, Marco Fiorucci

Figure 1 for Implicit neural representation for change detection
Figure 2 for Implicit neural representation for change detection
Figure 3 for Implicit neural representation for change detection
Figure 4 for Implicit neural representation for change detection

Detecting changes that occurred in a pair of 3D airborne LiDAR point clouds, acquired at two different times over the same geographical area, is a challenging task because of unmatching spatial supports and acquisition system noise. Most recent attempts to detect changes on point clouds are based on supervised methods, which require large labelled data unavailable in real-world applications. To address these issues, we propose an unsupervised approach that comprises two components: Neural Field (NF) for continuous shape reconstruction and a Gaussian Mixture Model for categorising changes. NF offer a grid-agnostic representation to encode bi-temporal point clouds with unmatched spatial support that can be regularised to increase high-frequency details and reduce noise. The reconstructions at each timestamp are compared at arbitrary spatial scales, leading to a significant increase in detection capabilities. We apply our method to a benchmark dataset of simulated LiDAR point clouds for urban sprawling. The dataset offers different challenging scenarios with different resolutions, input modalities and noise levels, allowing a multi-scenario comparison of our method with the current state-of-the-art. We boast the previous methods on this dataset by a 10% margin in intersection over union metric. In addition, we apply our methods to a real-world scenario to identify illegal excavation (looting) of archaeological sites and confirm that they match findings from field experts.

* Main article is 10 pages + 3 pages of supplementary. Conference style paper 
Viaarxiv icon

Beyond Exponential Graph: Communication-Efficient Topologies for Decentralized Learning via Finite-time Convergence

May 19, 2023
Yuki Takezawa, Ryoma Sato, Han Bao, Kenta Niwa, Makoto Yamada

Figure 1 for Beyond Exponential Graph: Communication-Efficient Topologies for Decentralized Learning via Finite-time Convergence
Figure 2 for Beyond Exponential Graph: Communication-Efficient Topologies for Decentralized Learning via Finite-time Convergence
Figure 3 for Beyond Exponential Graph: Communication-Efficient Topologies for Decentralized Learning via Finite-time Convergence
Figure 4 for Beyond Exponential Graph: Communication-Efficient Topologies for Decentralized Learning via Finite-time Convergence

Decentralized learning has recently been attracting increasing attention for its applications in parallel computation and privacy preservation. Many recent studies stated that the underlying network topology with a faster consensus rate (a.k.a. spectral gap) leads to a better convergence rate and accuracy for decentralized learning. However, a topology with a fast consensus rate, e.g., the exponential graph, generally has a large maximum degree, which incurs significant communication costs. Thus, seeking topologies with both a fast consensus rate and small maximum degree is important. In this study, we propose a novel topology combining both a fast consensus rate and small maximum degree called the Base-$(k + 1)$ Graph. Unlike the existing topologies, the Base-$(k + 1)$ Graph enables all nodes to reach the exact consensus after a finite number of iterations for any number of nodes and maximum degree k. Thanks to this favorable property, the Base-$(k + 1)$ Graph endows Decentralized SGD (DSGD) with both a faster convergence rate and more communication efficiency than the exponential graph. We conducted experiments with various topologies, demonstrating that the Base-$(k + 1)$ Graph enables various decentralized learning methods to achieve higher accuracy with better communication efficiency than the existing topologies.

Viaarxiv icon

Nystrom Method for Accurate and Scalable Implicit Differentiation

Feb 20, 2023
Ryuichiro Hataya, Makoto Yamada

Figure 1 for Nystrom Method for Accurate and Scalable Implicit Differentiation
Figure 2 for Nystrom Method for Accurate and Scalable Implicit Differentiation
Figure 3 for Nystrom Method for Accurate and Scalable Implicit Differentiation
Figure 4 for Nystrom Method for Accurate and Scalable Implicit Differentiation

The essential difficulty of gradient-based bilevel optimization using implicit differentiation is to estimate the inverse Hessian vector product with respect to neural network parameters. This paper proposes to tackle this problem by the Nystrom method and the Woodbury matrix identity, exploiting the low-rankness of the Hessian. Compared to existing methods using iterative approximation, such as conjugate gradient and the Neumann series approximation, the proposed method avoids numerical instability and can be efficiently computed in matrix operations without iterations. As a result, the proposed method works stably in various tasks and is faster than iterative approximations. Throughout experiments including large-scale hyperparameter optimization and meta learning, we demonstrate that the Nystrom method consistently achieves comparable or even superior performance to other approaches. The source code is available from https://github.com/moskomule/hypergrad.

* AISTATS 2023 
Viaarxiv icon

Optimal Transport for Change Detection on LiDAR Point Clouds

Feb 14, 2023
Marco Fiorucci, Peter Naylor, Makoto Yamada

Figure 1 for Optimal Transport for Change Detection on LiDAR Point Clouds
Figure 2 for Optimal Transport for Change Detection on LiDAR Point Clouds

The detection of changes occurring in multi-temporal remote sensing data plays a crucial role in monitoring several aspects of real life, such as disasters, deforestation, and urban planning. In the latter context, identifying both newly built and demolished buildings is essential to help landscape and city managers to promote sustainable development. While the use of airborne LiDAR point clouds has become widespread in urban change detection, the most common approaches require the transformation of a point cloud into a regular grid of interpolated height measurements, i.e. Digital Elevation Model (DEM). However, the DEM's interpolation step causes an information loss related to the height of the objects, affecting the detection capability of building changes, where the high resolution of LiDAR point clouds in the third dimension would be the most beneficial. Notwithstanding recent attempts to detect changes directly on point clouds using either a distance-based computation method or a semantic segmentation pre-processing step, only the M3C2 distance computation-based approach can identify both positive and negative changes, which is of paramount importance in urban planning. Motivated by the previous arguments, we introduce a principled change detection pipeline, based on optimal transport, capable of distinguishing between newly built buildings (positive changes) and demolished ones (negative changes). In this work, we propose to use unbalanced optimal transport to cope with the creation and destruction of mass related to building changes occurring in a bi-temporal pair of LiDAR point clouds. We demonstrate the efficacy of our approach on the only publicly available airborne LiDAR dataset for change detection by showing superior performance over the M3C2 and the previous optimal transport-based method presented by Nicolas Courty et al.at IGARSS 2016.

* Submitted to IEEE International Geoscience and Remote Sensing Symposium 2023 (IGARSS 2023) 
Viaarxiv icon

Structural Explanations for Graph Neural Networks using HSIC

Feb 04, 2023
Ayato Toyokuni, Makoto Yamada

Figure 1 for Structural Explanations for Graph Neural Networks using HSIC
Figure 2 for Structural Explanations for Graph Neural Networks using HSIC
Figure 3 for Structural Explanations for Graph Neural Networks using HSIC
Figure 4 for Structural Explanations for Graph Neural Networks using HSIC

Graph neural networks (GNNs) are a type of neural model that tackle graphical tasks in an end-to-end manner. Recently, GNNs have been receiving increased attention in machine learning and data mining communities because of the higher performance they achieve in various tasks, including graph classification, link prediction, and recommendation. However, the complicated dynamics of GNNs make it difficult to understand which parts of the graph features contribute more strongly to the predictions. To handle the interpretability issues, recently, various GNN explanation methods have been proposed. In this study, a flexible model agnostic explanation method is proposed to detect significant structures in graphs using the Hilbert-Schmidt independence criterion (HSIC), which captures the nonlinear dependency between two variables through kernels. More specifically, we extend the GraphLIME method for node explanation with a group lasso and a fused lasso-based node explanation method. The group and fused regularization with GraphLIME enables the interpretation of GNNs in substructure units. Then, we show that the proposed approach can be used for the explanation of sequential graph classification tasks. Through experiments, it is demonstrated that our method can identify crucial structures in a target graph in various settings.

Viaarxiv icon

Momentum Tracking: Momentum Acceleration for Decentralized Deep Learning on Heterogeneous Data

Sep 30, 2022
Yuki Takezawa, Han Bao, Kenta Niwa, Ryoma Sato, Makoto Yamada

Figure 1 for Momentum Tracking: Momentum Acceleration for Decentralized Deep Learning on Heterogeneous Data
Figure 2 for Momentum Tracking: Momentum Acceleration for Decentralized Deep Learning on Heterogeneous Data
Figure 3 for Momentum Tracking: Momentum Acceleration for Decentralized Deep Learning on Heterogeneous Data
Figure 4 for Momentum Tracking: Momentum Acceleration for Decentralized Deep Learning on Heterogeneous Data

SGD with momentum acceleration is one of the key components for improving the performance of neural networks. For decentralized learning, a straightforward approach using momentum acceleration is Distributed SGD (DSGD) with momentum acceleration (DSGDm). However, DSGDm performs worse than DSGD when the data distributions are statistically heterogeneous. Recently, several studies have addressed this issue and proposed methods with momentum acceleration that are more robust to data heterogeneity than DSGDm, although their convergence rates remain dependent on data heterogeneity and decrease when the data distributions are heterogeneous. In this study, we propose Momentum Tracking, which is a method with momentum acceleration whose convergence rate is proven to be independent of data heterogeneity. More specifically, we analyze the convergence rate of Momentum Tracking in the standard deep learning setting, where the objective function is non-convex and the stochastic gradient is used. Then, we identify that it is independent of data heterogeneity for any momentum coefficient $\beta\in [0, 1)$. Through image classification tasks, we demonstrate that Momentum Tracking is more robust to data heterogeneity than the existing decentralized learning methods with momentum acceleration and can consistently outperform these existing methods when the data distributions are heterogeneous.

Viaarxiv icon

Twin Papers: A Simple Framework of Causal Inference for Citations via Coupling

Aug 21, 2022
Ryoma Sato, Makoto Yamada, Hisashi Kashima

Figure 1 for Twin Papers: A Simple Framework of Causal Inference for Citations via Coupling
Figure 2 for Twin Papers: A Simple Framework of Causal Inference for Citations via Coupling
Figure 3 for Twin Papers: A Simple Framework of Causal Inference for Citations via Coupling
Figure 4 for Twin Papers: A Simple Framework of Causal Inference for Citations via Coupling

The research process includes many decisions, e.g., how to entitle and where to publish the paper. In this paper, we introduce a general framework for investigating the effects of such decisions. The main difficulty in investigating the effects is that we need to know counterfactual results, which are not available in reality. The key insight of our framework is inspired by the existing counterfactual analysis using twins, where the researchers regard twins as counterfactual units. The proposed framework regards a pair of papers that cite each other as twins. Such papers tend to be parallel works, on similar topics, and in similar communities. We investigate twin papers that adopted different decisions, observe the progress of the research impact brought by these studies, and estimate the effect of decisions by the difference in the impacts of these studies. We release our code and data, which we believe are highly beneficial owing to the scarcity of the dataset on counterfactual studies.

* CIKM 2022 short paper 
Viaarxiv icon

Inflating 2D Convolution Weights for Efficient Generation of 3D Medical Images

Aug 08, 2022
Yanbin Liu, Girish Dwivedi, Farid Boussaid, Frank Sanfilippo, Makoto Yamada, Mohammed Bennamoun

Figure 1 for Inflating 2D Convolution Weights for Efficient Generation of 3D Medical Images
Figure 2 for Inflating 2D Convolution Weights for Efficient Generation of 3D Medical Images
Figure 3 for Inflating 2D Convolution Weights for Efficient Generation of 3D Medical Images
Figure 4 for Inflating 2D Convolution Weights for Efficient Generation of 3D Medical Images

The generation of three-dimensional (3D) medical images can have great application potential since it takes into account the 3D anatomical structure. There are two problems, however, that prevent effective training of a 3D medical generative model: (1) 3D medical images are very expensive to acquire and annotate, resulting in an insufficient number of training images, (2) a large number of parameters are involved in 3D convolution. To address both problems, we propose a novel GAN model called 3D Split&Shuffle-GAN. In order to address the 3D data scarcity issue, we first pre-train a two-dimensional (2D) GAN model using abundant image slices and inflate the 2D convolution weights to improve initialization of the 3D GAN. Novel 3D network architectures are proposed for both the generator and discriminator of the GAN model to significantly reduce the number of parameters while maintaining the quality of image generation. A number of weight inflation strategies and parameter-efficient 3D architectures are investigated. Experiments on both heart (Stanford AIMI Coronary Calcium) and brain (Alzheimer's Disease Neuroimaging Initiative) datasets demonstrate that the proposed approach leads to improved 3D images generation quality with significantly fewer parameters.

* 10 pages 
Viaarxiv icon

Scale dependant layer for self-supervised nuclei encoding

Jul 22, 2022
Peter Naylor, Yao-Hung Hubert Tsai, Marick Laé, Makoto Yamada

Figure 1 for Scale dependant layer for self-supervised nuclei encoding
Figure 2 for Scale dependant layer for self-supervised nuclei encoding
Figure 3 for Scale dependant layer for self-supervised nuclei encoding
Figure 4 for Scale dependant layer for self-supervised nuclei encoding

Recent developments in self-supervised learning give us the possibility to further reduce human intervention in multi-step pipelines where the focus evolves around particular objects of interest. In the present paper, the focus lays in the nuclei in histopathology images. In particular we aim at extracting cellular information in an unsupervised manner for a downstream task. As nuclei present themselves in a variety of sizes, we propose a new Scale-dependant convolutional layer to bypass scaling issues when resizing nuclei. On three nuclei datasets, we benchmark the following methods: handcrafted, pre-trained ResNet, supervised ResNet and self-supervised features. We show that the proposed convolution layer boosts performance and that this layer combined with Barlows-Twins allows for better nuclei encoding compared to the supervised paradigm in the low sample setting and outperforms all other proposed unsupervised methods. In addition, we extend the existing TNBC dataset to incorporate nuclei class annotation in order to enrich and publicly release a small sample setting dataset for nuclei segmentation and classification.

* 13 pages, 6 figures, 2 tables 
Viaarxiv icon

Approximating 1-Wasserstein Distance with Trees

Jun 24, 2022
Makoto Yamada, Yuki Takezawa, Ryoma Sato, Han Bao, Zornitsa Kozareva, Sujith Ravi

Figure 1 for Approximating 1-Wasserstein Distance with Trees
Figure 2 for Approximating 1-Wasserstein Distance with Trees
Figure 3 for Approximating 1-Wasserstein Distance with Trees
Figure 4 for Approximating 1-Wasserstein Distance with Trees

Wasserstein distance, which measures the discrepancy between distributions, shows efficacy in various types of natural language processing (NLP) and computer vision (CV) applications. One of the challenges in estimating Wasserstein distance is that it is computationally expensive and does not scale well for many distribution comparison tasks. In this paper, we aim to approximate the 1-Wasserstein distance by the tree-Wasserstein distance (TWD), where TWD is a 1-Wasserstein distance with tree-based embedding and can be computed in linear time with respect to the number of nodes on a tree. More specifically, we propose a simple yet efficient L1-regularized approach to learning the weights of the edges in a tree. To this end, we first show that the 1-Wasserstein approximation problem can be formulated as a distance approximation problem using the shortest path distance on a tree. We then show that the shortest path distance can be represented by a linear model and can be formulated as a Lasso-based regression problem. Owing to the convex formulation, we can obtain a globally optimal solution efficiently. Moreover, we propose a tree-sliced variant of these methods. Through experiments, we demonstrated that the weighted TWD can accurately approximate the original 1-Wasserstein distance.

Viaarxiv icon