Search

Over the past few years, diffusion models have emerged as novel AI solutions, generating diverse multi-modal outputs from text prompts. Despite their capabilities, they face challenges in computing, such as excessive latency and energy consumption due to their iterative architecture. Although prior works specialized in transformer acceleration can be applied, the iterative nature of diffusion models remains unresolved. In this paper, we present EXION, the first SW-HW co-designed diffusion accelerator that solves the computation challenges by exploiting the unique inter- and intra-iteration output sparsity in diffusion models. To this end, we propose two SW-level optimizations. First, we introduce the FFN-Reuse algorithm that identifies and skips redundant computations in FFN layers across different iterations (inter-iteration sparsity). Second, we use a modified eager prediction method that employs two-step leading-one detection to accurately predict the attention score, skipping unnecessary computations within an iteration (intra-iteration sparsity). We also introduce a novel data compaction mechanism named ConMerge, which can enhance HW utilization by condensing and merging sparse matrices into compact forms. Finally, it has a dedicated HW architecture that supports the above sparsity-inducing algorithms, translating high output sparsity into improved energy efficiency and performance. To verify the feasibility of the EXION, we first demonstrate that it has no impact on accuracy in various types of multi-modal diffusion models. We then instantiate EXION in both server- and edge-level settings and compare its performance against GPUs with similar specifications. Our evaluation shows that EXION achieves dramatic improvements in performance and energy efficiency by 3.2-379.3x and 45.1-3067.6x compared to a server GPU and by 42.6-1090.9x and 196.9-4668.2x compared to an edge GPU., Comment: To appear in 2025 IEEE International Symposium on High-Performance Computer Architecture (HPCA 2025)

The goal of voice conversion is to transform the speech of a source speaker to sound like that of a reference speaker while preserving the original content. A key challenge is to extract disentangled linguistic content from the source and voice style from the reference. While existing approaches leverage various methods to isolate the two, a generalization still requires further attention, especially for robustness in zero-shot scenarios. In this paper, we achieve successful disentanglement of content and speaker features by tuning self-supervised speech features with adapters. The adapters are trained to dynamically encode nuanced features from rich self-supervised features, and the decoder fuses them to produce speech that accurately resembles the reference with minimal loss of content. Moreover, we leverage a conditional flow matching decoder with cross-attention speaker conditioning to further boost the synthesis quality and efficiency. Subjective and objective evaluations in a zero-shot scenario demonstrate that the proposed method outperforms existing models in speech quality and similarity to the reference speech., Comment: ICASSP 2025; demo available https://mm.kaist.ac.kr/projects/AdaptVC

The goal of this work is to generate natural speech in multiple languages while maintaining the same speaker identity, a task known as cross-lingual speech synthesis. A key challenge of cross-lingual speech synthesis is the language-speaker entanglement problem, which causes the quality of cross-lingual systems to lag behind that of intra-lingual systems. In this paper, we propose CrossSpeech++, which effectively disentangles language and speaker information and significantly improves the quality of cross-lingual speech synthesis. To this end, we break the complex speech generation pipeline into two simple components: language-dependent and speaker-dependent generators. The language-dependent generator produces linguistic variations that are not biased by specific speaker attributes. The speaker-dependent generator models acoustic variations that characterize speaker identity. By handling each type of information in separate modules, our method can effectively disentangle language and speaker representation. We conduct extensive experiments using various metrics, and demonstrate that CrossSpeech++ achieves significant improvements in cross-lingual speech synthesis, outperforming existing methods by a large margin.

Using the Subaru/FOCAS IFU capability, we examine the spatially resolved relationships between gas-phase metallicity, stellar mass, and star-formation rate surface densities (Sigma_* and Sigma_SFR, respectively) in extremely metal-poor galaxies (EMPGs) in the local universe. Our analysis includes 24 EMPGs, comprising 9,177 spaxels, which span a unique parameter space of local metallicity (12+log(O/H) = 6.9 to 7.9) and stellar mass surface density (Sigma_* ~ 10^5 to 10^7 Msun/kpc^2), extending beyond the range of existing large integral-field spectroscopic surveys. Through spatially resolved emission line diagnostics based on the [NII] BPT-diagram, we verify the absence of evolved active galactic nuclei in these EMPGs. Our findings reveal that, while the resolved mass-metallicity relation exhibits significant scatter in the low-mass regime, this scatter is closely correlated with local star-formation surface density. Specifically, metallicity decreases as Sigma_SFR increases for a given Sigma_*. Notably, half of the EMPGs show a distinct metal-poor horizontal branch on the resolved mass-metallicity relation. This feature typically appears at the peak clump with the highest Sigma_* and Sigma_SFR and is surrounded by a relatively metal-enriched ambient region. These findings support a scenario in which metal-poor gas infall fuels episodic star formation in EMPGs, consistent with the kinematic properties observed in these systems. In addition, we identify four EMPGs with exceptionally low central metallicities (12+log(O/H) <~ 7.2), which display only a metal-poor clump without a surrounding metal-rich region. This suggests that such ultra-low metallicity EMPGs, at less than a few percent of the solar metallicity, may serve as valuable analogs for galaxies in the early stages of galaxy evolution., Comment: 29 pages, 14 figures. Submitted to ApJ

In this paper, we introduce V2SFlow, a novel Video-to-Speech (V2S) framework designed to generate natural and intelligible speech directly from silent talking face videos. While recent V2S systems have shown promising results on constrained datasets with limited speakers and vocabularies, their performance often degrades on real-world, unconstrained datasets due to the inherent variability and complexity of speech signals. To address these challenges, we decompose the speech signal into manageable subspaces (content, pitch, and speaker information), each representing distinct speech attributes, and predict them directly from the visual input. To generate coherent and realistic speech from these predicted attributes, we employ a rectified flow matching decoder built on a Transformer architecture, which models efficient probabilistic pathways from random noise to the target speech distribution. Extensive experiments demonstrate that V2SFlow significantly outperforms state-of-the-art methods, even surpassing the naturalness of ground truth utterances.

The goal of this paper is to accelerate codec-based speech synthesis systems with minimum sacrifice to speech quality. We propose an enhanced inference method that allows for flexible trade-offs between speed and quality during inference without requiring additional training. Our core idea is to predict multiple tokens per inference step of the AR module using multiple prediction heads, resulting in a linear reduction in synthesis time as the number of heads increases. Furthermore, we introduce a novel speculative decoding technique that utilises a Viterbi-based algorithm to select the optimal sequence of generated tokens at each decoding step. In our experiments, we demonstrate that the time required to predict each token is reduced by a factor of 4 to 5 compared to baseline models, with minimal quality trade-off or even improvement in terms of speech intelligibility. Audio samples are available at: multpletokensprediction.github.io/multipletokensprediction.github.io/., Comment: Submitted to IEEE ICASSP 2025

This paper introduces SpoofCeleb, a dataset designed for Speech Deepfake Detection (SDD) and Spoofing-robust Automatic Speaker Verification (SASV), utilizing source data from real-world conditions and spoofing attacks generated by Text-To-Speech (TTS) systems also trained on the same real-world data. Robust recognition systems require speech data recorded in varied acoustic environments with different levels of noise to be trained. However, existing datasets typically include clean, high-quality recordings (bona fide data) due to the requirements for TTS training; studio-quality or well-recorded read speech is typically necessary to train TTS models. Existing SDD datasets also have limited usefulness for training SASV models due to insufficient speaker diversity. We present SpoofCeleb, which leverages a fully automated pipeline that processes the VoxCeleb1 dataset, transforming it into a suitable form for TTS training. We subsequently train 23 contemporary TTS systems. The resulting SpoofCeleb dataset comprises over 2.5 million utterances from 1,251 unique speakers, collected under natural, real-world conditions. The dataset includes carefully partitioned training, validation, and evaluation sets with well-controlled experimental protocols. We provide baseline results for both SDD and SASV tasks. All data, protocols, and baselines are publicly available at https://jungjee.github.io/spoofceleb., Comment: 9 pages, 2 figures, 8 tables

Text-to-speech (TTS) systems are traditionally trained using modest databases of studio-quality, prompted or read speech collected in benign acoustic environments such as anechoic rooms. The recent literature nonetheless shows efforts to train TTS systems using data collected in the wild. While this approach allows for the use of massive quantities of natural speech, until now, there are no common datasets. We introduce the TTS In the Wild (TITW) dataset, the result of a fully automated pipeline, in this case, applied to the VoxCeleb1 dataset commonly used for speaker recognition. We further propose two training sets. TITW-Hard is derived from the transcription, segmentation, and selection of VoxCeleb1 source data. TITW-Easy is derived from the additional application of enhancement and additional data selection based on DNSMOS. We show that a number of recent TTS models can be trained successfully using TITW-Easy, but that it remains extremely challenging to produce similar results using TITW-Hard. Both the dataset and protocols are publicly available and support the benchmarking of TTS systems trained using TITW data., Comment: 5 pages, submitted to ICASSP 2025 as a conference paper

We present a proof-of-concept simulation-based inference on $\Omega_{\rm m}$ and $\sigma_{8}$ from the SDSS BOSS LOWZ NGC catalog using neural networks and domain generalization techniques without the need of summary statistics. Using rapid lightcone simulations, ${\rm L{\scriptsize -PICOLA}}$, mock galaxy catalogs are produced that fully incorporate the observational effects. The collection of galaxies is fed as input to a point cloud-based network, ${\texttt{Minkowski-PointNet}}$. We also add relatively more accurate ${\rm G{\scriptsize ADGET}}$ mocks to obtain robust and generalizable neural networks. By explicitly learning the representations which reduces the discrepancies between the two different datasets via the semantic alignment loss term, we show that the latent space configuration aligns into a single plane in which the two cosmological parameters form clear axes. Consequently, during inference, the SDSS BOSS LOWZ NGC catalog maps onto the plane, demonstrating effective generalization and improving prediction accuracy compared to non-generalized models. Results from the ensemble of 25 independently trained machines find $\Omega_{\rm m}=0.339 \pm 0.056$ and $\sigma_{8}=0.801 \pm 0.061$, inferred only from the distribution of galaxies in the lightcone slices without relying on any indirect summary statistics. A single machine that best adapts to the ${\rm G{\scriptsize ADGET}}$ mocks yields a tighter prediction of $\Omega_{\rm m}=0.282 \pm 0.014$ and $\sigma_{8}=0.786 \pm 0.036$. We emphasize that adaptation across multiple domains can enhance the robustness of the neural networks in observational data., Comment: 28 pages, 14 figures; Accepted for publication in ApJ

We introduce a GPU-accelerated hybrid hydro/N-body code (Enzo-N) designed to address the challenges of concurrently simulating star clusters and their parent galaxies. This task has been exceedingly challenging, primarily due to the considerable computational time required, which stems from the substantial scale difference between galaxies (~ 0.1 Mpc) and star clusters (~ pc). Yet, this significant scale separation means that particles within star clusters perceive those outside the star cluster in a semi-stationary state. By leveraging this aspect, we integrate the direct N-body code (Nbody6++GPU) into the cosmological (magneto-)hydrodynamic code (Enzo) through the utilization of the semi-stationary background acceleration approximation. We solve the dynamics of particles within star clusters using the direct N-body solver with regularization for few-body interactions, while evolving particles outside -- dark matter, gas, and stars -- using the particle-mesh gravity solver and hydrodynamic methods. We demonstrate that Enzo-N successfully simulates the co-evolution of star clusters and their parent galaxies, capturing phenomena such as core collapse of the star cluster and tidal stripping due to galactic tides. This comprehensive framework opens up new possibilities for studying the evolution of star clusters within galaxies, offering insights that were previously inaccessible., Comment: 15 pages, 13 figures, Accepted for publication in ApJ

The AGORA Cosmorun (arXiv:2106.09738) is a set of hydrodynamical cosmological zoom-in simulations carried out within the AGORA High-resolution Galaxy Simulations Comparison Project (arXiv:1308.2669,arXiv:1610.03066). These simulations show the formation and evolution of a Milky Way-sized galaxy using eight of the most widely used numerical codes in the community (Art-I, Enzo, Ramses, Changa, Gadget-3, Gear, Gizmo, and Arepo). In this short report, we describe the public release of the raw output data from all of these simulations at z = 8, 7, 6, 5, 4, 3, 2 (plus at z=1, 0 when available), and several metadata files containing the halo centers, virial quantities, and merger trees. The data from even thinner timesteps will be released as soon as the upcoming collaboration papers (VII-IX) are submitted and accepted., Comment: ArXiv Data Release from the AGORA Collaboration. arXiv admin note: text overlap with arXiv:2001.04354

This paper introduces VoxSim, a dataset of perceptual voice similarity ratings. Recent efforts to automate the assessment of speech synthesis technologies have primarily focused on predicting mean opinion score of naturalness, leaving speaker voice similarity relatively unexplored due to a lack of extensive training data. To address this, we generate about 41k utterance pairs from the VoxCeleb dataset, a widely utilised speech dataset for speaker recognition, and collect nearly 70k speaker similarity scores through a listening test. VoxSim offers a valuable resource for the development and benchmarking of speaker similarity prediction models. We provide baseline results of speaker similarity prediction models on the VoxSim test set and further demonstrate that the model trained on our dataset generalises to the out-of-domain VCC2018 dataset., Comment: INTERSPEECH 2024. The dataset is available from https://mm.kaist.ac.kr/projects/voxsim/

We introduce a novel software called TCSpy which is designed to efficiently control a multi-telescope array through network-based protocols. The primary objectives of TCSpy include centralized control of the array, support for diverse observation modes, and swift responses to the follow-up observations of astronomical transients. To achieve these objectives, TCSpy utilizes the ASCOM Alpaca protocol in conjunction with Alpyca, establishing robust communication among multiple telescope units. For the practical application of TCSpy, we implement TCSpy within the 7-Dimensional Telescope (7DT). 7DT is a telescope array consisting of 20, 0.5-m telescopes, equipped with 40 different medium-band filters. The main scientific goals of 7DT include detecting the optical counterparts of gravitational-wave sources, identifying kilonovae, and the spectral mapping of the southern sky. Through the integration of TCSpy, 7DT can achieve these scientific objectives with its unique observation modes and rapid follow-up capabilities., Comment: 8 pages, 5 figures, SPIE Astronomical Telescopes + Instrumentation 2024

Center for the Gravitational-Wave Universe at Seoul National University has been operating its main observational facility, the 7-Dimensional Telescope (7DT) since October 2023. Located at El Sauce Observatory in Chilean Rio Hurtado Valley, 7DT consists of 20 50-cm telescopes equipped with 40 medium-band filters of 25 nm full width at half maximum along with a CMOS camera of 61 megapixels. 7DT produces about 1 TB per night of spectral mapping image data including calibration, and the byproduct of the data reduction pipeline once our planned three layered surveys (Reference Imaging Survey, Wide Field Survey, and Intensive Monitoring Survey) start in 2024. We are expecting to generate 1 PB per year by combining raw data, reduced data, and data products (e.g. calibrated stacked images, spectral cubes, and object catalogs). To incorporate this huge amount of data, we now have a data storage for 1 PB which we will increment by 1 PB per year. We also have a high-performance computation facility that is equipped with 2 NVIDIA A100 GPU cards since we plan to carry out real-time data reduction and analysis for follow-up observation data of gravitational wave events. To incorporate this, we established a 400 Mbps network connection between the facilities in Korea and Chile. Taking advantage of the high-performance network, we have been carrying out fully remote operations since October 2023. In this talk, we present details of designing, planning, and executing the ground-based telescope facility project, especially within low-budget academic environments. While we cover as much ground as possible, we will emphasize human resource management, project risk management, and financial contingency management., Comment: 7 pages, 1 figures, Proceedings of the SPIE conference "Modeling, Systems Engineering, and Project Management for Astronomy XI" SPIE Astronomical Telescopes + Instrumentation 2024 (Paper No. 13099-77)

The 7-Dimensional Telescope (7DT) is a multi-telescope system designed to identify electromagnetic (EM) counterparts of gravitational-wave (GW) sources. Consisting of 20 50-cm telescopes along with 40 medium-band filters of 25 nm width, 7DT can obtain spectral mapping images for a large field of view (~1.25 square degrees). Along with flexible operation, real-time data reduction, and analysis, the 7DT's spectral mapping capability enables 7DT to follow up GW events quickly and discover EM counterparts. Among 20 planned telescopes, 12 units are deployed at the El Sauce Observatory located at Rio Hurtado Valley in Chile. Since we obtained the first light of 7DT in October 2023, we started its commissioning procedures including examination of bias levels, master flat production, and spectrophotometric standardization. In this talk, we present 7DT instruments and their set-up, commissioning procedures, and data characteristics of 7DT along with our three-layered surveys which are assumed to be initiated in early 2024., Comment: 11 pages, 3 figures, Proceedings of the SPIE conference 13094 Ground-based and Airborne Telescope X, SPIE Astronomical Telescopes + Instrumentations 2024 (Paper No. 13094-034)

This work proposes an efficient method to enhance the quality of corrupted speech signals by leveraging both acoustic and visual cues. While existing diffusion-based approaches have demonstrated remarkable quality, their applicability is limited by slow inference speeds and computational complexity. To address this issue, we present FlowAVSE which enhances the inference speed and reduces the number of learnable parameters without degrading the output quality. In particular, we employ a conditional flow matching algorithm that enables the generation of high-quality speech in a single sampling step. Moreover, we increase efficiency by optimizing the underlying U-net architecture of diffusion-based systems. Our experiments demonstrate that FlowAVSE achieves 22 times faster inference speed and reduces the model size by half while maintaining the output quality. The demo page is available at: https://cyongong.github.io/FlowAVSE.github.io/, Comment: INTERSPEECH 2024

The goal of this work is to simultaneously generate natural talking faces and speech outputs from text. We achieve this by integrating Talking Face Generation (TFG) and Text-to-Speech (TTS) systems into a unified framework. We address the main challenges of each task: (1) generating a range of head poses representative of real-world scenarios, and (2) ensuring voice consistency despite variations in facial motion for the same identity. To tackle these issues, we introduce a motion sampler based on conditional flow matching, which is capable of high-quality motion code generation in an efficient way. Moreover, we introduce a novel conditioning method for the TTS system, which utilises motion-removed features from the TFG model to yield uniform speech outputs. Our extensive experiments demonstrate that our method effectively creates natural-looking talking faces and speech that accurately match the input text. To our knowledge, this is the first effort to build a multimodal synthesis system that can generalise to unseen identities., Comment: CVPR 2024

In this fourth paper from the AGORA Collaboration, we study the evolution down to redshift $z=2$ and below of a set of cosmological zoom-in simulations of a Milky Way mass galaxy by eight of the leading hydrodynamic simulation codes. We also compare this CosmoRun suite of simulations with dark matter-only simulations by the same eight codes. We analyze general properties of the halo and galaxy at $z=4$ and 3, and before the last major merger, focusing on the formation of well-defined rotationally-supported disks, the mass-metallicity relation, the specific star formation rate, the gas metallicity gradients, and the non-axisymmetric structures in the stellar disks. Codes generally converge well to the stellar-to-halo mass ratios predicted by semi-analytic models at $z\sim$2. We see that almost all the hydro codes develop rotationally-supported structures at low redshifts. Most agree within 0.5 dex with the observed MZR at high and intermediate redshifts, and reproduce the gas metallicity gradients obtained from analytical models and low-redshift observations. We confirm that the inter-code differences in the halo assembly history reported in the first paper of the collaboration also exist in CosmoRun, making the code-to-code comparison more difficult. We show that such differences are mainly due to variations in code-dependent parameters that control the time-stepping strategy of the gravity solver. We find that variations in the early stellar feedback can also result in differences in the timing of the low-redshift mergers. All the simulation data down to $z=2$ and the auxiliary data will be made publicly available., Comment: Submitted to ApJ. Visit the AGORA Collaboration website (www.agorasimulations.org ) for more information

We analyze and compare the satellite halo populations at $z\sim2$ in the high-resolution cosmological zoom-in simulations of a $10^{12}\,{\rm M}_{\odot}$ target halo ($z=0$ mass) carried out on eight widely-used astrophysical simulation codes ({\sc Art-I}, {\sc Enzo}, {\sc Ramses}, {\sc Changa}, {\sc Gadget-3}, {\sc Gear}, {\sc Arepo-t}, and {\sc Gizmo}) for the {\it AGORA} High-resolution Galaxy Simulations Comparison Project. We use slightly different redshift epochs near $z=2$ for each code (hereafter ``$z\sim2$') at which the eight simulations are in the same stage in the target halo's merger history. After identifying the matched pairs of halos between the {\it CosmoRun} simulations and the DMO simulations, we discover that each {\it CosmoRun} halo tends to be less massive than its DMO counterpart. When we consider only the halos containing stellar particles at $z\sim2$, the number of satellite {\it galaxies} is significantly fewer than that of dark matter halos in all participating {\it AGORA} simulations, and is comparable to the number of present-day satellites near the Milky Way or M31. The so-called ``missing satellite problem' is fully resolved across all participating codes simply by implementing the common baryonic physics adopted in {\it AGORA} and the stellar feedback prescription commonly used in each code, with sufficient numerical resolution ($\lesssim100$ proper pc at $z=2$). We also compare other properties such as the stellar mass$-$halo mass relation and the mass$-$metallicity relation. Our work highlights the value of comparison studies such as {\it AGORA}, where outstanding problems in galaxy formation theory are studied simultaneously on multiple numerical platforms., Comment: Accepted for publication to the ApJ, 19 pages, 10 figures. Visit the AGORA Collaboration website (www.agorasimulations.org) for more information. For summary video, please see https://www.youtube.com/watch?v=HShRJYvPe1k

We analyze the circumgalactic medium (CGM) for eight commonly-used cosmological codes in the AGORA collaboration. The codes are calibrated to use identical initial conditions, cosmology, heating and cooling, and star formation thresholds, but each evolves with its own unique code architecture and stellar feedback implementation. Here, we analyze the results of these simulations in terms of the structure, composition, and phase dynamics of the CGM. We show properties such as metal distribution, ionization levels, and kinematics are effective tracers of the effects of the different code feedback and implementation methods, and as such they can be highly divergent between simulations. This is merely a fiducial set of models, against which we will in the future compare multiple feedback recipes for each code. Nevertheless, we find that the large parameter space these simulations establish can help disentangle the different variables that affect observable quantities in the CGM, e.g., showing that abundances for ions with higher ionization energy are more strongly determined by the simulation's metallicity, while abundances for ions with lower ionization energy are more strongly determined by the gas density and temperature., Comment: 32 pages, 17 figures. Published in ApJ

Dark matter-deficient galaxies (DMDGs) discovered in the survey of ultra-diffuse galaxies (UDGs), in apparent conflict with standard CDM, may be produced by high-velocity galaxy-galaxy collisions, the $\textit{Mini-bullet}$ scenario. Recent observations of an aligned trail of $7-11$ UDGs near NGC1052, including DMDGs DF2 and DF4, suggesting a common formation event, $\sim8.9\pm1.5$ Gyr ago, provide a test. Hydro/N-body simulations, supplemented by galaxy orbit integrations, demonstrate that satellite-satellite collisions outside the host-galaxy virial radius can reproduce the observed UDGs in the NGC1052 group. A trail of $\sim10$ DMDGs is shown to form, including two massive ones that replicate the observed motions of DF2 and DF4. The linear relation, $v=Ax+v_{0}$, conjectured previously to relate positions ($x$) and velocities ($v$) of the aligned DMDGs as a signature of the collision event, is approximately obeyed, but individual DMDGs can deviate significantly from it. The progenitors whose collision spawned the trail of DMDGs survive the collision without, themselves, becoming DMDGs. We predict one progenitor is located at the end of the trail, testable by observing the difference between its stars, formed pre-collision, from those of the DMDGs, formed post-collision. By contrast, stellar ages and metallicities of the DMDGs are nearly identical. We further offer a hint that the tidal field of host NGC1052 may contribute to making DMDGs diffuse. $\Lambda$CDM simulation in a 100 cMpc box finds our required initial conditions $\sim10$ times at $z<3$. These results indicate current observations are consistent with the $\textit{Mini-bullet}$ scenario., Comment: 27 pages, 14 figures, minor revisions for clarification only in response to referee report, + 2 new figures. For associated movies and information, see https://joohyun-lee.github.io/research/DMDG3/

We introduce a novel halo/galaxy matching technique between two cosmological simulations with different resolutions, which utilizes the positions and masses of halos along their subhalo merger tree. With this tool, we conduct a study of resolution biases through the {\it galaxy-by-galaxy} inspection of a pair of simulations that have the same simulation configuration but different mass resolutions, utilizing a suite of {\sc IllustrisTNG} simulations to assess the impact on galaxy properties. We find that, with the subgrid physics model calibrated for TNG100-1, subhalos in TNG100-1 (high resolution) have $\lesssim0.5$ dex higher stellar masses than their counterparts in the TNG100-2 (low-resolution). It is also discovered that the subhalos with $M_{\mathrm{gas}}\sim10^{8.5}\,{\rm M}_\odot$ in TNG100-1 have $\sim0.5$ dex higher gas mass than those in TNG100-2. The mass profiles of the subhalos reveal that the dark matter masses of subhalos in TNG100-2 converge well with those from TNG100-1, except within 4 kpc of the resolution limit. The differences in stellar mass and hot gas mass are most pronounced in the central region. We exploit machine learning to build a correction mapping for the physical quantities of subhalos from low- to high-resolution simulations (TNG300-1 and TNG100-1), which enables us to find an efficient way to compile a high-resolution galaxy catalog even from a low-resolution simulation. Our tools can easily be applied to other large cosmological simulations, testing and mitigating the resolution biases of their numerical codes and subgrid physics models., Comment: 24 pages, 14 figures; Accepted for publication in to ApJ

The goal of this work is to reconstruct high quality speech from lip motions alone, a task also known as lip-to-speech. A key challenge of lip-to-speech systems is the one-to-many mapping caused by (1) the existence of homophenes and (2) multiple speech variations, resulting in a mispronounced and over-smoothed speech. In this paper, we propose a novel lip-to-speech system that significantly improves the generation quality by alleviating the one-to-many mapping problem from multiple perspectives. Specifically, we incorporate (1) self-supervised speech representations to disambiguate homophenes, and (2) acoustic variance information to model diverse speech styles. Additionally, to better solve the aforementioned problem, we employ a flow based post-net which captures and refines the details of the generated speech. We perform extensive experiments on two datasets, and demonstrate that our method achieves the generation quality close to that of real human utterance, outperforming existing methods in terms of speech naturalness and intelligibility by a large margin. Synthesised samples are available at our demo page: https://mm.kaist.ac.kr/projects/LTBS., Comment: Accepted to AAAI 2024

Recent abrupt changes of CW Leonis may indicate that we are witnessing the moment that the central carbon star is evolving off the Asymptotic Giant Branch (AGB) and entering into the pre-planetary nebula (PPN) phase. The recent appearance of a red compact peak at the predicted stellar position is possibly an unveiling event of the star, and the radial beams emerging from the stellar position resemble the feature of the PPN Egg Nebula. The increase of light curve over two decades is also extraordinary, and it is possibly related to the phase transition. Decadal-period variations are further found in the residuals of light curves, in the relative brightness of radial beams, and in the extended halo brightness distribution. Further monitoring of the recent dramatic and decadal-scale changes of this most well-known carbon star CW Leonis at the tip of AGB is still highly essential, and will help us gain a more concrete understanding on the conditions for transition between the late stellar evolutionary phases., Comment: 6 pages, 4 figures, Winds of Stars and Exoplanets. Proceedings of the International Astronomical Union, Volume 370, pp. 78-83

For the first time in the world, we succeeded in synthesizing the room-temperature superconductor ($T_c \ge 400$ K, 127$^\circ$C) working at ambient pressure with a modified lead-apatite (LK-99) structure. The superconductivity of LK-99 is proved with the Critical temperature ($T_c$), Zero-resistivity, Critical current ($I_c$), Critical magnetic field ($H_c$), and the Meissner effect. The superconductivity of LK-99 originates from minute structural distortion by a slight volume shrinkage (0.48 %), not by external factors such as temperature and pressure. The shrinkage is caused by Cu$^{2+}$ substitution of Pb$^{2+}$(2) ions in the insulating network of Pb(2)-phosphate and it generates the stress. It concurrently transfers to Pb(1) of the cylindrical column resulting in distortion of the cylindrical column interface, which creates superconducting quantum wells (SQWs) in the interface. The heat capacity results indicated that the new model is suitable for explaining the superconductivity of LK-99. The unique structure of LK-99 that allows the minute distorted structure to be maintained in the interfaces is the most important factor that LK-99 maintains and exhibits superconductivity at room temperatures and ambient pressure.

We present demography of the dynamics and gas-mass fraction of 33 extremely metal-poor galaxies (EMPGs) with metallicities of $0.015-0.195~Z_\odot$ and low stellar masses of $10^4-10^8~M_\odot$ in the local universe. We conduct deep optical integral-field spectroscopy (IFS) for the low-mass EMPGs with the medium high resolution ($R=7500$) grism of the 8m-Subaru FOCAS IFU instrument by the EMPRESS 3D survey, and investigate H$\alpha$ emission of the EMPGs. Exploiting the resolution high enough for the low-mass galaxies, we derive gas dynamics with the H$\alpha$ lines by the fitting of 3-dimensional disk models. We obtain an average maximum rotation velocity ($v_\mathrm{rot}$) of $15\pm3~\mathrm{km~s^{-1}}$ and an average intrinsic velocity dispersion ($\sigma_0$) of $27\pm10~\mathrm{km~s^{-1}}$ for 15 spatially resolved EMPGs out of the 33 EMPGs, and find that all of the 15 EMPGs have $v_\mathrm{rot}/\sigma_0<1$ suggesting dispersion dominated systems. There is a clear decreasing trend of $v_\mathrm{rot}/\sigma_0$ with the decreasing stellar mass and metallicity. We derive the gas mass fraction ($f_\mathrm{gas}$) for all of the 33 EMPGs, and find no clear dependence on stellar mass and metallicity. These $v_\mathrm{rot}/\sigma_0$ and $f_\mathrm{gas}$ trends should be compared with young high-$z$ galaxies observed by the forthcoming JWST IFS programs to understand the physical origins of the EMPGs in the local universe., Comment: 18 pages, 9 figures, accepted for publication in ApJ

While recent text-to-speech (TTS) systems have made remarkable strides toward human-level quality, the performance of cross-lingual TTS lags behind that of intra-lingual TTS. This gap is mainly rooted from the speaker-language entanglement problem in cross-lingual TTS. In this paper, we propose CrossSpeech which improves the quality of cross-lingual speech by effectively disentangling speaker and language information in the level of acoustic feature space. Specifically, CrossSpeech decomposes the speech generation pipeline into the speaker-independent generator (SIG) and speaker-dependent generator (SDG). The SIG produces the speaker-independent acoustic representation which is not biased to specific speaker distributions. On the other hand, the SDG models speaker-dependent speech variation that characterizes speaker attributes. By handling each information separately, CrossSpeech can obtain disentangled speaker and language representations. From the experiments, we verify that CrossSpeech achieves significant improvements in cross-lingual TTS, especially in terms of speaker similarity to the target speaker., Comment: Accepted to ICASSP 2023

We search for local extremely metal-poor galaxies (EMPGs), selecting photometric candidates by broadband color excess and machine-learning techniques with the SDSS photometric data. After removing stellar contaminants by shallow spectroscopy with Seimei and Nayuta telescopes, we confirm that three candidates are EMPGs with 0.05--0.1 $Z_\odot$ by deep Magellan/MagE spectroscopy for faint {\sc[Oiii]}$\lambda$4363 lines. Using a statistical sample consisting of 105 spectroscopically-confirmed EMPGs taken from our study and the literature, we calculate cross-correlation function (CCF) of the EMPGs and all SDSS galaxies to quantify environments of EMPGs. Comparing another CCF of all SDSS galaxies and comparison SDSS galaxies in the same stellar mass range ($10^{7.0}-10^{8.4} M_\odot$), we find no significant ($>1\sigma$) difference between these two CCFs. We also compare mass-metallicity relations (MZRs) of the EMPGs and those of galaxies at $z\sim$ 0--4 with a steady chemical evolution model and find that the EMPG MZR is comparable with the model prediction on average. These clustering and chemical properties of EMPGs are explained by a scenario of stochastic metal-poor gas accretion on metal-rich galaxies showing metal-poor star formation. Extending the broadband color-excess technique to a high-$z$ EMPG search, we select 17 candidates of $z\sim$ 4--5 EMPGs with the deep ($\simeq30$ mag) near-infrared JWST/NIRCam images obtained by ERO and ERS programs. We find galaxy candidates with negligible {\sc[Oiii]}$\lambda\lambda$4959,5007 emission weaker than the local EMPGs and known high-$z$ galaxies, suggesting that some of these candidates may fall in 0--0.01 $Z_\odot$, which potentially break the lowest metallicity limit known to date.