Search

In the image classification task, deep neural networks frequently rely on bias attributes that are spuriously correlated with a target class in the presence of dataset bias, resulting in degraded performance when applied to data without bias attributes. The task of debiasing aims to compel classifiers to learn intrinsic attributes that inherently define a target class rather than focusing on bias attributes. While recent approaches mainly focus on emphasizing the learning of data samples without bias attributes (i.e., bias-conflicting samples) compared to samples with bias attributes (i.e., bias-aligned samples), they fall short of directly guiding models where to focus for learning intrinsic features. To address this limitation, this paper proposes a method that provides the model with explicit spatial guidance that indicates the region of intrinsic features. We first identify the intrinsic features by investigating the class-discerning common features between a bias-aligned (BA) sample and a bias-conflicting (BC) sample (i.e., bias-contrastive pair). Next, we enhance the intrinsic features in the BA sample that are relatively under-exploited for prediction compared to the BC sample. To construct the bias-contrastive pair without using bias information, we introduce a bias-negative score that distinguishes BC samples from BA samples employing a biased model. The experiments demonstrate that our method achieves state-of-the-art performance on synthetic and real-world datasets with various levels of bias severity., Comment: Accepted to CVPR 2024

We present a black hole effect by strategically leveraging a conformal mapping in elastic continuum with a curved-space framework, which is less stringent compared to a Schwarzschild model transformed to isotropic refractive-index profiles. In the conformal map approach, the two-dimensional point singularity associated to the black hole effect is accomplished by physical plates with near-to-zero thickness. The analog gravity around the singularity results in highly confined energy and lagged timings within a branch cut of the conformal map. These effects are quantified both numerically and experimentally in reference to control trials in which the thickness is not modulated. The findings would deepen our understanding of the elastic analog in mimicking gravitational phenomena, as well as establish the elastic continuum framework for developing a generic design recipe in the presence of the index singularity. Geometric landscapes with elastically curved surfaces would be applicable in a variety of applications such as sensing, imaging, vibration isolation, and energy harvesting. © 2022 American Physical Society.

In image classification, "debiasing" aims to train a classifier to be less susceptible to dataset bias, the strong correlation between peripheral attributes of data samples and a target class. For example, even if the frog class in the dataset mainly consists of frog images with a swamp background (i.e., bias-aligned samples), a debiased classifier should be able to correctly classify a frog at a beach (i.e., bias-conflicting samples). Recent debiasing approaches commonly use two components for debiasing, a biased model $f_B$ and a debiased model $f_D$. $f_B$ is trained to focus on bias-aligned samples (i.e., overfitted to the bias) while $f_D$ is mainly trained with bias-conflicting samples by concentrating on samples which $f_B$ fails to learn, leading $f_D$ to be less susceptible to the dataset bias. While the state-of-the-art debiasing techniques have aimed to better train $f_D$, we focus on training $f_B$, an overlooked component until now. Our empirical analysis reveals that removing the bias-conflicting samples from the training set for $f_B$ is important for improving the debiasing performance of $f_D$. This is due to the fact that the bias-conflicting samples work as noisy samples for amplifying the bias for $f_B$ since those samples do not include the bias attribute. To this end, we propose a simple yet effective data sample selection method which removes the bias-conflicting samples to construct a bias-amplified dataset for training $f_B$. Our data sample selection method can be directly applied to existing reweighting-based debiasing approaches, obtaining consistent performance boost and achieving the state-of-the-art performance on both synthetic and real-world datasets., Comment: Accepted to AAAI 2023

The development of sound-absorbing materials for noise reduction in daily life has been a prolonged issue that also applies to a recognized need for submarine anechoic tiles to stay independent from SONAR (SOund NAvigation Ranging). Here, we present an underwater stealth metasurface that uses split-orifice-conduit (SOC) hybrid resonators to significantly reduce its acoustic reflectance. A theoretical analysis of SOC elements provides an approach to quantifying acoustic characteristics using the transfer matrix method in a single metasurface. The findings confirm that we can tune the absorption with respect to a resonating frequency by adjusting geometrical parameters. Utilizing a hybrid mechanism that enables easy access to coupled resonances, we obtain broadband absorption spectra even in the presence of a covariant sound speed profile in the deep sea and a thermoviscous effect on unit cells of the metasurface. Such a metasurface will provide a further step toward developing feasible underwater stealth technologies for submarines and remains to be experimentally demonstrated.

Transformation optics, which is generically applicable to other classical waves such as acoustic and elastic waves, provides an emerging design paradigm to manipulate waves. However, some lenses and optical-transformation devices require a singular refractive index; meeting this requirement is a significant challenge. A method called transmutation can relax some types of index singularity into finite anisotropy around the singularity. Here, we show that such lenses with a singularity for flexural waves can be obtained by approaching a near-zero thickness of the plate precisely at the location of the singularity. As examples, we demonstrate a series of Eaton lenses theoretically and experimentally by projecting the refractive index in space onto the thickness in plates and by working in a broad frequency range in which impedance mismatch is negligible. This framework offers an insight into feasible methods that can be used to develop singular devices such as cloaking devices on thin flexible curved plates and can be further extended to a general methodology for shaping elastic waves. We hope that this elastic platform can also be a test bed to indirectly study unprecedented phenomena enabled by gravitational and quantum fields in terms of analog models. © 2021 American Physical Society.

Transformation optics, which is generically applicable to other classical waves such as acoustic and elastic waves, provides an emerging design paradigm to manipulate waves. However, some lenses and optical-transformation devices require a singular refractive index; meeting this requirement is a significant challenge. A method called transmutation can relax some types of index singularity into finite anisotropy around the singularity. Here, we show that such lenses with a singularity for flexural waves can be obtained by approaching a near-zero thickness of the plate precisely at the location of the singularity. As examples, we demonstrate a series of Eaton lenses theoretically and experimentally by projecting the refractive index in space onto the thickness in plates and by working in a broad frequency range in which impedance mismatch is negligible. This framework offers an insight into feasible methods that can be used to develop singular devices such as cloaking devices on thin flexible curved plates and can be further extended to a general methodology for shaping elastic waves. We hope that this elastic platform can also be a test bed to indirectly study unprecedented phenomena enabled by gravitational and quantum fields in terms of analog models. © 2021 American Physical Society.

The development of sound-absorbing materials for noise reduction in daily life has been a prolonged issue that also applies to a recognized need for submarine anechoic tiles to stay independent from SONAR (SOund NAvigation Ranging). Here, we present an underwater stealth metasurface that uses split-orifice-conduit (SOC) hybrid resonators to significantly reduce its acoustic reflectance. A theoretical analysis of SOC elements provides an approach to quantifying acoustic characteristics using the transfer matrix method in a single metasurface. The findings confirm that we can tune the absorption with respect to a resonating frequency by adjusting geometrical parameters. Utilizing a hybrid mechanism that enables easy access to coupled resonances, we obtain broadband absorption spectra even in the presence of a covariant sound speed profile in the deep sea and a thermoviscous effect on unit cells of the metasurface. Such a metasurface will provide a further step toward developing feasible underwater stealth technologies for submarines and remains to be experimentally demonstrated.

Despite the impressive performance of deep networks in vision, language, and healthcare, unpredictable behaviors on samples from the distribution different than the training distribution cause severe problems in deployment. For better reliability of neural-network-based classifiers, we define a new task, natural attribute-based shift (NAS) detection, to detect the samples shifted from the training distribution by some natural attribute such as age of subjects or brightness of images. Using the natural attributes present in existing datasets, we introduce benchmark datasets in vision, language, and medical for NAS detection. Further, we conduct an extensive evaluation of prior representative out-of-distribution (OOD) detection methods on NAS datasets and observe an inconsistency in their performance. To understand this, we provide an analysis on the relationship between the location of NAS samples in the feature space and the performance of distance- and confidence-based OOD detection methods. Based on the analysis, we split NAS samples into three categories and further suggest a simple modification to the training objective to obtain an improved OOD detection method that is capable of detecting samples from all NAS categories.

The development of sound-absorbing materials for noise reduction in daily life has been a prolonged issue that also applies to a recognized need for submarine anechoic tiles to stay independent from SONAR (SOund NAvigation Ranging). Here, we present an underwater stealth metasurface that uses split-orifice-conduit (SOC) hybrid resonators to significantly reduce its acoustic reflectance. A theoretical analysis of SOC elements provides an approach to quantifying acoustic characteristics using the transfer matrix method in a single metasurface. The findings confirm that we can tune the absorption with respect to a resonating frequency by adjusting geometrical parameters. Utilizing a hybrid mechanism that enables easy access to coupled resonances, we obtain broadband absorption spectra even in the presence of a covariant sound speed profile in the deep sea and a thermoviscous effect on unit cells of the metasurface. Such a metasurface will provide a further step toward developing feasible underwater stealth technologies for submarines and remains to be experimentally demonstrated.

Transformation optics, which is generically applicable to other classical waves such as acoustic and elastic waves, provides an emerging design paradigm to manipulate waves. However, some lenses and optical-transformation devices require a singular refractive index; meeting this requirement is a significant challenge. A method called transmutation can relax some types of index singularity into finite anisotropy around the singularity. Here, we show that such lenses with a singularity for flexural waves can be obtained by approaching a near-zero thickness of the plate precisely at the location of the singularity. As examples, we demonstrate a series of Eaton lenses theoretically and experimentally by projecting the refractive index in space onto the thickness in plates and by working in a broad frequency range in which impedance mismatch is negligible. This framework offers an insight into feasible methods that can be used to develop singular devices such as cloaking devices on thin flexible curved plates and can be further extended to a general methodology for shaping elastic waves. We hope that this elastic platform can also be a test bed to indirectly study unprecedented phenomena enabled by gravitational and quantum fields in terms of analog models. © 2021 American Physical Society.

Deep neural networks for automatic image colorization often suffer from the color-bleeding artifact, a problematic color spreading near the boundaries between adjacent objects. Such color-bleeding artifacts debase the reality of generated outputs, limiting the applicability of colorization models in practice. Although previous approaches have attempted to address this problem in an automatic manner, they tend to work only in limited cases where a high contrast of gray-scale values are given in an input image. Alternatively, leveraging user interactions would be a promising approach for solving this color-breeding artifacts. In this paper, we propose a novel edge-enhancing network for the regions of interest via simple user scribbles indicating where to enhance. In addition, our method requires a minimal amount of effort from users for their satisfactory enhancement. Experimental results demonstrate that our interactive edge-enhancing approach effectively improves the color-bleeding artifacts compared to the existing baselines across various datasets., Comment: Accepted to ICCV 2021 as oral presentation (3% acceptance rate)

We evaluate the out-of-distribution (OOD) detection performance of self-supervised learning (SSL) techniques with a new evaluation framework. Unlike the previous evaluation methods, the proposed framework adjusts the distance of OOD samples from the in-distribution samples. We evaluate an extensive combination of OOD detection algorithms on three different implementations of the proposed framework using simulated samples, images, and text. SSL methods consistently demonstrated the improved OOD detection performance in all evaluation settings., Comment: Accepted to NeurIPS 2020 Workshop: Self-Supervised Learning - Theory and Practice

Joan Flasch Artists' Book Collection

A new methodology to measure coded image/video quality using the just-noticeable-difference (JND) idea was proposed. Several small JND-based image/video quality datasets were released by the Media Communications Lab at the University of Southern California. In this work, we present an effort to build a large-scale JND-based coded video quality dataset. The dataset consists of 220 5-second sequences in four resolutions (i.e., $1920 \times 1080$, $1280 \times 720$, $960 \times 540$ and $640 \times 360$). For each of the 880 video clips, we encode it using the H.264 codec with $QP=1, \cdots, 51$ and measure the first three JND points with 30+ subjects. The dataset is called the "VideoSet", which is an acronym for "Video Subject Evaluation Test (SET)". This work describes the subjective test procedure, detection and removal of outlying measured data, and the properties of collected JND data. Finally, the significance and implications of the VideoSet to future video coding research and standardization efforts are pointed out. All source/coded video clips as well as measured JND data included in the VideoSet are available to the public in the IEEE DataPort.

In this paper, we explicitly evaluate the one-shot quantum non-signalling assisted zero-error classical capacities $\M_0^{\mathrm{QNS}}$ for qubit channels. In particular, we show that for nonunital qubit channels, $\M_0^{\mathrm{QNS}}=1$, which implies that in the one-shot setting, nonunital qubit channels cannot transmit any information with zero probability of error even when assisted by quantum non-signalling correlations. Furthermore, we show that for qubit channels, $\M_0^{\mathrm{QNS}}$ equals to the one-shot entanglement-assisted zero-error classical capacities. This means that for a single use of a qubit channel, quantum non-signalling correlations are not more powerful than shared entanglement., Comment: 6 pages, no figure