Whenever the concept of high-performance cloth simulation is brought up, GPU acceleration is almost always the first that comes to mind. Leveraging immense parallelization, GPU algorithms have demonstrated significant success recently, whereas CPU methods are somewhat overlooked. Indeed, the need for an efficient CPU simulator is evident and pressing. In many scenarios, high-end GPUs may be unavailable or are already allocated to other tasks, such as rendering and shading. A high-performance CPU alternative can greatly boost the overall system capability and user experience. Inspired by this demand, this paper proposes a CPU algorithm for high-resolution cloth simulation. By partitioning the garment model into multiple (but not massive) sub-meshes or domains, we assign per-domain computations to individual CPU processors. Borrowing the idea of projective dynamics that breaks the computation into global and local steps, our key contribution is a new parallelization paradigm at domains for both global and local steps so that domain-level calculations are sequential and lightweight. The CPU has much fewer processing units than a GPU. Our algorithm mitigates this disadvantage by wisely balancing the scale of the parallelization and convergence. We validate our method in a wide range of simulation problems involving high-resolution garment models. Performance-wise, our method is at least one order faster than existing CPU methods, and it delivers a similar performance compared with the state-of-the-art GPU algorithms in many examples, but without using a GPU.

Reproducing realistic collective behaviors presents a captivating yet formidable challenge. Traditional rule-based methods rely on hand-crafted principles, limiting motion diversity and realism in generated collective behaviors. Recent imitation learning methods learn from data but often require ground-truth motion trajectories and struggle with authenticity, especially in high-density groups with erratic movements. In this paper, we present a scalable approach, Collective Behavior Imitation Learning (CBIL), for learning fish schooling behavior directly from videos , without relying on captured motion trajectories. Our method first leverages Video Representation Learning, in which a Masked Video AutoEncoder (MVAE) extracts implicit states from video inputs in a self-supervised manner. The MVAE effectively maps 2D observations to implicit states that are compact and expressive for following the imitation learning stage. Then, we propose a novel adversarial imitation learning method to effectively capture complex movements of the schools of fish, enabling efficient imitation of the distribution of motion patterns measured in the latent space. It also incorporates bio-inspired rewards alongside priors to regularize and stabilize training. Once trained, CBIL can be used for various animation tasks with the learned collective motion priors. We further show its effectiveness across different species. Finally, we demonstrate the application of our system in detecting abnormal fish behavior from in-the-wild videos.

Realtime speech denoising has been long studied. Almost all existing methods process the incoming data stream using a sliding window of fixed-size. Yet, we show that the use of fixed-size sliding window may lead to an accumulating lag, especially in presence of other background computing processes that may occupy CPU resources. In response, we propose a new sliding window strategy and a lightweight neural network to leverage it. Our experiments show that the proposed approach achieves denoising quality on a par with the stateof-the-art realtime denoising models. More importantly, our approach is faster, maintaining a stable realtime performance even when the available computing power fluctuates.

Our goal is to synthesize realistic underwater scenes with various fish species in different fish cages, which can be utilized to train computer vision models to automate fish counting and sizing tasks. It is a challenging problem to prepare a sufficiently diverse labeled dataset of images from aquatic environments. We solve this challenge by introducing an adaptive bio-inspired fish simulation. The behavior of caged fish changes based on the species, size and number of fish, and the size and shape of the cage, among other variables. However, a method to autonomously achieve schooling behavior for caged fish did not exist. In this paper, we propose a method for achieving schooling behavior for any given combination of variables, using multi-agent deep reinforcement learning (DRL) in various fish cages in arbitrary environments. Furthermore, to visually reproduce the underwater scene in different locations and seasons, we incorporate a physically-based underwater simulation.

We present a bio-inspired fish simulation platform, which we call "Foids", to generate realistic synthetic datasets for an use in computer vision algorithm training. This is a first-of-its-kind synthetic dataset platform for fish, which generates all the 3D scenes just with a simulation. One of the major challenges in deep learning based computer vision is the preparation of the annotated dataset. It is already hard to collect a good quality video dataset with enough variations; moreover, it is a painful process to annotate a sufficiently large video dataset frame by frame. This is especially true when it comes to a fish dataset because it is difficult to set up a camera underwater and the number of fish (target objects) in the scene can range up to 30,000 in a fish cage on a fish farm. All of these fish need to be annotated with labels such as a bounding box or silhouette, which can take hours to complete manually, even for only a few minutes of video. We solve this challenge by introducing a realistic synthetic dataset generation platform that incorporates details of biology and ecology studied in the aquaculture field. Because it is a simulated scene, it is easy to generate the scene data with annotation labels from the 3D mesh geometry data and transformation matrix. To this end, we develop an automated fish counting system utilizing the part of synthetic dataset that shows comparable counting accuracy to human eyes, which reduces the time compared to the manual process, and reduces physical injuries sustained by the fish.

We introduce a deep learning model for speech denoising, a long-standing challenge in audio analysis arising in numerous applications. Our approach is based on a key observation about human speech: there is often a short pause between each sentence or word. In a recorded speech signal, those pauses introduce a series of time periods during which only noise is present. We leverage these incidental silent intervals to learn a model for automatic speech denoising given only mono-channel audio. Detected silent intervals over time expose not just pure noise but its time-varying features, allowing the model to learn noise dynamics and suppress it from the speech signal. Experiments on multiple datasets confirm the pivotal role of silent interval detection for speech denoising, and our method outperforms several state-of-the-art denoising methods, including those that accept only audio input (like ours) and those that denoise based on audiovisual input (and hence require more information). We also show that our method enjoys excellent generalization properties, such as denoising spoken languages not seen during training.

We propose a new method called ”Dark Side Ternary Stars (DSTS)”, which is expected to allow emotional personal agents to interact with users. DSTS focuses on proper nouns to adapt preference to users. In this paper, we report only on ”GAGPL” in the DSTS. GAGPL is a method for calculating positive or negative value per sentence for an individual user. We aim to represent famous Japanese proverb ”The misery of others is as sweet as honey”. Our proposed method consists of Genetic Algorithms (GA) and Genetic Programming Like (GPL) algorithms. GA learns the positive or negative values of general word to individuals. GPL calculates the positive or negative value per sentence. We applied this method to the news article text data in Japanese professional baseball. As the experiment results, it shows that this method is effective.

The purpose of the article is the development of an interface which closely adapts to the individual. By quantifying the frustration as the human manipulates machines from a biomedical signal and making it into teaching signals of machine learning (ML), we aim at the development of a system in which the machine adapts to the human. The authors extract the characteristic vector of whether the examinee is in discomfort or not from electroencephalograms (EEG) and electromyograms (EMG). An artificial neural network (ANN) is employed to extract the vector. For the machine learning, reinforcement learning is used and the rewards are an extracted signal from physiological signals. As a basic experiment for extracting discomfort and comfort from the physiological signals, the EEG measurement experiment was carried out under an unpleasant sound environment for 20 examinees. The input signals to ANN for the characteristic vector extraction was examined. By affixing piezoelectric films on the eyebrow, the movement of the eyebrow was measured. Finally, the results of measuring EEG and EMG simultaneously under the situation in which frustration accumulated for the examinee are shown. We use reinforcement learning (RL) to control the behavior of the Khepera robot.