Search

Learning generalized models from biased data is an important undertaking toward fairness in deep learning. To address this issue, recent studies attempt to identify and leverage bias-conflicting samples free from spurious correlations without prior knowledge of bias or an unbiased set. However, spurious correlation remains an ongoing challenge, primarily due to the difficulty in precisely detecting these samples. In this paper, inspired by the similarities between mislabeled samples and bias-conflicting samples, we approach this challenge from a novel perspective of mislabeled sample detection. Specifically, we delve into Influence Function, one of the standard methods for mislabeled sample detection, for identifying bias-conflicting samples and propose a simple yet effective remedy for biased models by leveraging them. Through comprehensive analysis and experiments on diverse datasets, we demonstrate that our new perspective can boost the precision of detection and rectify biased models effectively. Furthermore, our approach is complementary to existing methods, showing performance improvement even when applied to models that have already undergone recent debiasing techniques.

Verifying the violation of Bell's inequality is one of the most representative methods to demonstrate that entangled photon pairs prepared in a quantum optics-based system exhibit quantum properties. While experiments on Bell inequality violations have been theoretically well-established and extensively conducted to implement various quantum information technologies in laboratory settings, mathematical modeling for accurately predicting the distribution of high-intensity entangled photon pairs in high-loss environments remains an issue that requires further research. As the brightness of the entangled photon pairs increases, the influence of multi-photon effects becomes more significant, leading to a decrease in the CHSH value $S$ and also a reduction in the standard deviation of the CHSH value $\Delta S$. Therefore, a new analysis of the $(S-2)/\Delta S$ value is required to more precisely confirm the degree of CHSH inequality violation including the reliability of $S$. In this paper, we propose a mathematical model to predict the $(S-2)/\Delta S$ value as a function of the brightness of the entangled photon pair source, and we also suggest the need to optimize the brightness of this source. Additionally, we provide experimental evidence supporting this model. The experiment confirms that when the mean photon number is $\mu=0.026$ in an entanglement distribution setup with a total loss of $-19.03$ dB, the CHSH value drops to 2.69, while the $(S-2)/\Delta S$ value increases to 60.95., Comment: 11 pages, 1 table, and 4 figures

In this study, we introduces a parameter-efficient model that outperforms traditional models in time series forecasting, by integrating High-order Polynomial Projection (HiPPO) theory into the Kolmogorov-Arnold network (KAN) framework. This HiPPO-KAN model achieves superior performance on long sequence data without increasing parameter count. Experimental results demonstrate that HiPPO-KAN maintains a constant parameter count while varying window sizes and prediction horizons, in contrast to KAN, whose parameter count increases linearly with window size. Surprisingly, although the HiPPO-KAN model keeps a constant parameter count as increasing window size, it significantly outperforms KAN model at larger window sizes. These results indicate that HiPPO-KAN offers significant parameter efficiency and scalability advantages for time series forecasting. Additionally, we address the lagging problem commonly encountered in time series forecasting models, where predictions fail to promptly capture sudden changes in the data. We achieve this by modifying the loss function to compute the MSE directly on the coefficient vectors in the HiPPO domain. This adjustment effectively resolves the lagging problem, resulting in predictions that closely follow the actual time series data. By incorporating HiPPO theory into KAN, this study showcases an efficient approach for handling long sequences with improved predictive accuracy, offering practical contributions for applications in large-scale time series data., Comment: 16 pages, 6 figures, 2 tables

Large language models (LLMs) have demonstrated strong capabilities across various language tasks, notably through instruction-tuning methods. However, LLMs face challenges in visualizing complex, real-world data through charts and plots. Firstly, existing datasets rarely cover a full range of chart types, such as 3D, volumetric, and gridded charts. Secondly, supervised fine-tuning methods do not fully leverage the intricate relationships within rich datasets, including text, code, and figures. To address these challenges, we propose a hierarchical pipeline and a new dataset for chart generation. Our dataset, Text2Chart31, includes 31 unique plot types referring to the Matplotlib library, with 11.1K tuples of descriptions, code, data tables, and plots. Moreover, we introduce a reinforcement learning-based instruction tuning technique for chart generation tasks without requiring human feedback. Our experiments show that this approach significantly enhances the model performance, enabling smaller models to outperform larger open-source models and be comparable to state-of-the-art proprietary models in data visualization tasks. We make the code and dataset available at https://github.com/fatemehpesaran310/Text2Chart31., Comment: EMNLP 2024 Main. Code and dataset are released at https://github.com/fatemehpesaran310/Text2Chart31

Clinical experts employ diverse philosophies and strategies in patient care, influenced by regional patient populations. However, existing medical artificial intelligence (AI) models are often trained on data distributions that disproportionately reflect highly prevalent patterns, reinforcing biases and overlooking the diverse expertise of clinicians. To overcome this limitation, we introduce the Mixture of Multicenter Experts (MoME) approach. This method strategically integrates specialized expertise from diverse clinical strategies, enhancing the AI model's ability to generalize and adapt across multiple medical centers. The MoME-based multimodal target volume delineation model, trained with few-shot samples including images and clinical notes from each medical center, outperformed baseline methods in prostate cancer radiotherapy target delineation. The advantages of MoME were most pronounced when data characteristics varied across centers or when data availability was limited, demonstrating its potential for broader clinical applications. Therefore, the MoME framework enables the deployment of AI-based target volume delineation models in resource-constrained medical facilities by adapting to specific preferences of each medical center only using a few sample data, without the need for data sharing between institutions. Expanding the number of multicenter experts within the MoME framework will significantly enhance the generalizability, while also improving the usability and adaptability of clinical AI applications in the field of precision radiation oncology., Comment: 39 pages

Beta-amyloid positron emission tomography (A$\beta$-PET) imaging has become a critical tool in Alzheimer's disease (AD) research and diagnosis, providing insights into the pathological accumulation of amyloid plaques, one of the hallmarks of AD. However, the high cost, limited availability, and exposure to radioactivity restrict the widespread use of A$\beta$-PET imaging, leading to a scarcity of comprehensive datasets. Previous studies have suggested that structural magnetic resonance imaging (MRI), which is more readily available, may serve as a viable alternative for synthesizing A$\beta$-PET images. In this study, we propose an approach to utilize 3D diffusion models to synthesize A$\beta$-PET images from T1-weighted MRI scans, aiming to overcome the limitations associated with direct PET imaging. Our method generates high-quality A$\beta$-PET images for cognitive normal cases, although it is less effective for mild cognitive impairment (MCI) patients due to the variability in A$\beta$ deposition patterns among subjects. Our preliminary results suggest that incorporating additional data, such as a larger sample of MCI cases and multi-modality information including clinical and demographic details, cognitive and functional assessments, and longitudinal data, may be necessary to improve A$\beta$-PET image synthesis for MCI patients.

While large language models (LLMs) exhibit significant utility across various domains, they simultaneously are susceptible to exploitation for unethical purposes, including academic misconduct and dissemination of misinformation. Consequently, AI-generated text detection systems have emerged as a countermeasure. However, these detection mechanisms demonstrate vulnerability to evasion techniques and lack robustness against textual manipulations. This paper introduces back-translation as a novel technique for evading detection, underscoring the need to enhance the robustness of current detection systems. The proposed method involves translating AI-generated text through multiple languages before back-translating to English. We present a model that combines these back-translated texts to produce a manipulated version of the original AI-generated text. Our findings demonstrate that the manipulated text retains the original semantics while significantly reducing the true positive rate (TPR) of existing detection methods. We evaluate this technique on nine AI detectors, including six open-source and three proprietary systems, revealing their susceptibility to back-translation manipulation. In response to the identified shortcomings of existing AI text detectors, we present a countermeasure to improve the robustness against this form of manipulation. Our results indicate that the TPR of the proposed method declines by only 1.85% after back-translation manipulation. Furthermore, we build a large dataset of 720k texts using eight different LLMs. Our dataset contains both human-authored and LLM-generated texts in various domains and writing styles to assess the performance of our method and existing detectors. This dataset is publicly shared for the benefit of the research community.

In radiation therapy (RT), the reliance on pre-treatment computed tomography (CT) images encounter challenges due to anatomical changes, necessitating adaptive planning. Daily cone-beam CT (CBCT) imaging, pivotal for therapy adjustment, falls short in tissue density accuracy. To address this, our innovative approach integrates diffusion models for CT image generation, offering precise control over data synthesis. Leveraging a self-training method with knowledge distillation, we maximize CBCT data during therapy, complemented by sparse paired fan-beam CTs. This strategy, incorporated into state-of-the-art diffusion-based models, surpasses conventional methods like Pix2pix and CycleGAN. A meticulously curated dataset of 2800 paired CBCT and CT scans, supplemented by 4200 CBCT scans, undergoes preprocessing and teacher model training, including the Brownian Bridge Diffusion Model (BBDM). Pseudo-label CT images are generated, resulting in a dataset combining 5600 CT images with corresponding CBCT images. Thorough evaluation using MSE, SSIM, PSNR and LPIPS demonstrates superior performance against Pix2pix and CycleGAN. Our approach shows promise in generating high-quality CT images from CBCT scans in RT., Comment: MICCAI 2024

Recent advancements in automatic speaker verification (ASV) studies have been achieved by leveraging large-scale pretrained networks. In this study, we analyze the approaches toward such a paradigm and underline the significance of interlayer information processing as a result. Accordingly, we present a novel approach for exploiting the multilayered nature of pretrained models for ASV, which comprises a layer/frame-level network and two steps of pooling architectures for each layer and frame axis. Specifically, we let convolutional architecture directly processes a stack of layer outputs.Then, we present a channel attention-based scheme of gauging layer significance and squeeze the layer level with the most representative value. Finally, attentive statistics over frame-level representations yield a single vector speaker embedding. Comparative experiments are designed using versatile data environments and diverse pretraining models to validate the proposed approach. The experimental results demonstrate the stability of the approach using multi-layer outputs in leveraging pretrained architectures. Then, we verify the superiority of the proposed ASV backend structure, which involves layer-wise operations, in terms of performance improvement along with cost efficiency compared to the conventional method. The ablation study shows how the proposed interlayer processing aids in maximizing the advantage of utilizing pretrained models., Comment: Preprint

Exciton plasmon polaritons have gained increasing interests over recent years due to their versatile properties emerging by the underlying light-matter coupling and making them potential candidates for new photonic applications. We have advanced this concept by studying thin films of laterally aligned J-type aggregates of self-assembled tetra-bay phenoxy-dendronized perylene bisimide (PBI) molecules, arranged in a helical manner of three strains on a silver surface. As a result of the interaction between the uniformly aligned dipole moments and the surface plasmons of a thin silver layer underneath, the excitonic state at 1.94 eV evolves into dispersions in absorption and emission, both characterized by a distinct anisotropy. The coupling constant defined by the scalar product of the transition dipole moment $\vec{\mu}$ and the surface plasmon wavevector $\vec{k}_x$ shows a pronounced two-fold rotational symmetry with values between almost 0 to 28 meV. Complementary TD-DFT calculations of the angular dependent absorption and photoluminescence provide insights in the coherent energy exchange between the excitonic and plasmonic sub-systems. Additionally, power dependent PL studies yield first evidence that the diffusion length of the coupled exciton-plasmon polaritons exceeds that of the mere Frenkel state in neat PBI by at least one order of magnitude. Our results not only demonstrate the possibility to control the photo-physical properties of strongly coupled states by their spatially anisotropic light-matter interaction but also reveal innovative strategies to influence opto-electronic device operation by the directional transport of hybrid state energy., Comment: 26 pages, 7 figures, 2 tables, Supporting Information

A high-brightness entangled photon pair (HBEPP) source is essential for conducting entanglement-based quantum key distribution (QKD) between a satellite and a ground station. While an ultrabright source can overcome significant losses in satellite-based QKD (SQKD) and increase the sifted key rate, it also induces the multi-photon effect, raising the system's error rate. To accurately estimate system performance, we first present an analytical model for calculating the measurement probabilities of HBEPP distribution through an asymmetric loss channel. Based on this model, we propose the use of a passive-intensity HBEPP source for SQKD systems, assuming a polarization-independent channel and threshold detectors for measurement. We confirm that fixing the mean photon number at $\bar{\mu}=0.1$ achieves a performance of $99.7\%$ compared to the ideal one-way communication entanglement-based SQKD protocol, which is effectively optimizing the HBEPP source brightness in accordance with system losses., Comment: 9 pages, 5 figures, 3 tables

In this work, we discuss evaluating video foundation models in a fair and robust manner. Unlike language or image foundation models, many video foundation models are evaluated with differing parameters (such as sampling rate, number of frames, pretraining steps, etc.), making fair and robust comparisons challenging. Therefore, we present a carefully designed evaluation framework for measuring two core capabilities of video comprehension: appearance and motion understanding. Our findings reveal that existing video foundation models, whether text-supervised like UMT or InternVideo2, or self-supervised like V-JEPA, exhibit limitations in at least one of these capabilities. As an alternative, we introduce TWLV-I, a new video foundation model that constructs robust visual representations for both motion- and appearance-based videos. Based on the average top-1 accuracy of linear probing on five action recognition benchmarks, pretrained only on publicly accessible datasets, our model shows a 4.6%p improvement compared to V-JEPA (ViT-L) and a 7.7%p improvement compared to UMT (ViT-L). Even when compared to much larger models, our model demonstrates a 7.2%p improvement compared to DFN (ViT-H), a 2.7%p improvement compared to V-JEPA (ViT-H) and a 2.8%p improvement compared to InternVideo2 (ViT-g). We provide embedding vectors obtained by TWLV-I from videos of several commonly used video benchmarks, along with evaluation source code that can directly utilize these embeddings. The code is available at https://github.com/twelvelabs-io/video-embeddings-evaluation-framework., Comment: 17 pages; Twelve Labs Technical Report

Source-Free domain adaptive Object Detection (SFOD) is a promising strategy for deploying trained detectors to new, unlabeled domains without accessing source data, addressing significant concerns around data privacy and efficiency. Most SFOD methods leverage a Mean-Teacher (MT) self-training paradigm relying heavily on High-confidence Pseudo Labels (HPL). However, these HPL often overlook small instances that undergo significant appearance changes with domain shifts. Additionally, HPL ignore instances with low confidence due to the scarcity of training samples, resulting in biased adaptation toward familiar instances from the source domain. To address this limitation, we introduce the Low-confidence Pseudo Label Distillation (LPLD) loss within the Mean-Teacher based SFOD framework. This novel approach is designed to leverage the proposals from Region Proposal Network (RPN), which potentially encompasses hard-to-detect objects in unfamiliar domains. Initially, we extract HPL using a standard pseudo-labeling technique and mine a set of Low-confidence Pseudo Labels (LPL) from proposals generated by RPN, leaving those that do not overlap significantly with HPL. These LPL are further refined by leveraging class-relation information and reducing the effect of inherent noise for the LPLD loss calculation. Furthermore, we use feature distance to adaptively weight the LPLD loss to focus on LPL containing a larger foreground area. Our method outperforms previous SFOD methods on four cross-domain object detection benchmarks. Extensive experiments demonstrate that our LPLD loss leads to effective adaptation by reducing false negatives and facilitating the use of domain-invariant knowledge from the source model. Code is available at https://github.com/junia3/LPLD., Comment: ECCV 2024

With large language models (LLMs), conversational search engines shift how users retrieve information from the web by enabling natural conversations to express their search intents over multiple turns. Users' natural conversation embodies rich but implicit signals of users' search intents and evaluation of search results to understand user experience with the system. However, it is underexplored how and why users ask follow-up queries to continue conversations with conversational search engines and how the follow-up queries signal users' satisfaction. From qualitative analysis of 250 conversational turns from an in-lab user evaluation of Naver Cue:, a commercial conversational search engine, we propose a taxonomy of 18 users' follow-up query patterns from conversational search, comprising two major axes: (1) users' motivations behind continuing conversations (N = 7) and (2) actions of follow-up queries (N = 11). Compared to the existing literature on query reformulations, we uncovered a new set of motivations and actions behind follow-up queries, including asking for subjective opinions or providing natural language feedback on the engine's responses. To analyze conversational search logs with our taxonomy in a scalable and efficient manner, we built an LLM-powered classifier (73% accuracy). With our classifier, we analyzed 2,061 conversational tuples collected from real-world usage logs of Cue: and examined how the conversation patterns from our taxonomy correlates with satisfaction. Our initial findings suggest some signals of dissatisfactions, such as Clarifying Queries, Excluding Condition, and Substituting Condition with follow-up queries. We envision our approach could contribute to automated evaluation of conversation search experience by providing satisfaction signals and grounds for realistic user simulations., Comment: Accepted to LLM4Eval @ SIGIR 2024 - The First Workshop on Large Language Models (LLMs) for Evaluation in Information Retrieval

Recent advancements in text-to-image (T2I) models have greatly benefited from large-scale datasets, but they also pose significant risks due to the potential generation of unsafe content. To mitigate this issue, researchers have developed unlearning techniques to remove the model's ability to generate potentially harmful content. However, these methods are easily bypassed by adversarial attacks, making them unreliable for ensuring the safety of generated images. In this paper, we propose Direct Unlearning Optimization (DUO), a novel framework for removing Not Safe For Work (NSFW) content from T2I models while preserving their performance on unrelated topics. DUO employs a preference optimization approach using curated paired image data, ensuring that the model learns to remove unsafe visual concepts while retaining unrelated features. Furthermore, we introduce an output-preserving regularization term to maintain the model's generative capabilities on safe content. Extensive experiments demonstrate that DUO can robustly defend against various state-of-the-art red teaming methods without significant performance degradation on unrelated topics, as measured by FID and CLIP scores. Our work contributes to the development of safer and more reliable T2I models, paving the way for their responsible deployment in both closed-source and open-source scenarios., Comment: Extended abstract accepted in GenLaw 2024 workshop @ ICML2024

Expressive Text-to-Speech (TTS) using reference speech has been studied extensively to synthesize natural speech, but there are limitations to obtaining well-represented styles and improving model generalization ability. In this study, we present Diffusion-based EXpressive TTS (DEX-TTS), an acoustic model designed for reference-based speech synthesis with enhanced style representations. Based on a general diffusion TTS framework, DEX-TTS includes encoders and adapters to handle styles extracted from reference speech. Key innovations contain the differentiation of styles into time-invariant and time-variant categories for effective style extraction, as well as the design of encoders and adapters with high generalization ability. In addition, we introduce overlapping patchify and convolution-frequency patch embedding strategies to improve DiT-based diffusion networks for TTS. DEX-TTS yields outstanding performance in terms of objective and subjective evaluation in English multi-speaker and emotional multi-speaker datasets, without relying on pre-training strategies. Lastly, the comparison results for the general TTS on a single-speaker dataset verify the effectiveness of our enhanced diffusion backbone. Demos are available here., Comment: Preprint

Score distillation sampling (SDS), the methodology in which the score from pretrained 2D diffusion models is distilled into 3D representation, has recently brought significant advancements in text-to-3D generation task. However, this approach is still confronted with critical geometric inconsistency problems such as the Janus problem. Starting from a hypothesis that such inconsistency problems may be induced by multiview inconsistencies between 2D scores predicted from various viewpoints, we introduce GSD, a simple and general plug-and-play framework for incorporating 3D consistency and therefore geometry awareness into the SDS process. Our methodology is composed of three components: 3D consistent noising, designed to produce 3D consistent noise maps that perfectly follow the standard Gaussian distribution, geometry-based gradient warping for identifying correspondences between predicted gradients of different viewpoints, and novel gradient consistency loss to optimize the scene geometry toward producing more consistent gradients. We demonstrate that our method significantly improves performance, successfully addressing the geometric inconsistency problems in text-to-3D generation task with minimal computation cost and being compatible with existing score distillation-based models. Our project page is available at https://ku-cvlab.github.io/GSD/.

We propose a metasurface antenna capable of real time holographic beam steering. An array of reconfigurable dipoeles can generate on demand far field patterns of radiation through the specific encoding of meta atomic states. i.e., the configuration of each dipole. Suitable states for the generation of the desired patterns can be identified using iteartion, but this is very slow and needs to be done for each far field pattern. Here, we present a deep learning based method for the control of a metasurface antenna with point dipole elements that vary in their state using dipole polarizability. Instead of iteration, we adopt a deep learning algorithm that combines an autoencoder with an electromagnetic scattering equation to determin the states required for a target far field pattern in real time. The scattering equation from Born approximation is used as the decoder in training the neural network, and analytic Green's function calculation is used to check the validity of Born approximation. Our learning based algorithm requires a computing time of within in 200 microseconds to determine the meta atomic states, thus enabling the real time opeartion of a holographic antenna.

3D reconstruction from multi-view images is one of the fundamental challenges in computer vision and graphics. Recently, 3D Gaussian Splatting (3DGS) has emerged as a promising technique capable of real-time rendering with high-quality 3D reconstruction. This method utilizes 3D Gaussian representation and tile-based splatting techniques, bypassing the expensive neural field querying. Despite its potential, 3DGS encounters challenges, including needle-like artifacts, suboptimal geometries, and inaccurate normals, due to the Gaussians converging into anisotropic Gaussians with one dominant variance. We propose using effective rank analysis to examine the shape statistics of 3D Gaussian primitives, and identify the Gaussians indeed converge into needle-like shapes with the effective rank 1. To address this, we introduce effective rank as a regularization, which constrains the structure of the Gaussians. Our new regularization method enhances normal and geometry reconstruction while reducing needle-like artifacts. The approach can be integrated as an add-on module to other 3DGS variants, improving their quality without compromising visual fidelity., Comment: project page: https://junhahyung.github.io/erankgs.github.io

We examined archival Far Ultraviolet Spectroscopic Explorer data to search for far-ultraviolet emission lines in the starburst galaxy M82. The observations were made in an outflow region that extends beyond the galactic disk. We found the O VI $\lambda\lambda$ 1032, 1038 emission lines from the galaxy's southern outflow region. The O VI lines suggest that the outflowing warm-hot gas is undergoing radiative cooling. We measured a radial velocity of $\sim$420 km s$^{-1}$ from the O VI lines, which is faster than the velocity seen in H$\alpha$ observations. The O VI $\lambda$1038 emission line seems to be blended with the C II $\lambda$1037 emission line, which has a radial velocity of $\sim$300 km s$^{-1}$, similar to what is observed in H$\alpha$ observations. The outflow medium of M82 appears to be composed of gas in multiple phases with varying temperatures and kinematics. Future spectroscopic observations in high energy regimes covering a wider spatial area are necessary to understand better the properties of the warm-hot gas medium in the outflow., Comment: 9 pages, 5 figures, Accepted to AJ

Neural Radiance Fields (NeRF) have shown remarkable performance in learning 3D scenes. However, NeRF exhibits vulnerability when confronted with distractors in the training images -- unexpected objects are present only within specific views, such as moving entities like pedestrians or birds. Excluding distractors during dataset construction is a straightforward solution, but without prior knowledge of their types and quantities, it becomes prohibitively expensive. In this paper, we propose PruNeRF, a segment-centric dataset pruning framework via 3D spatial consistency, that effectively identifies and prunes the distractors. We first examine existing metrics for measuring pixel-wise distraction and introduce Influence Functions for more accurate measurements. Then, we assess 3D spatial consistency using a depth-based reprojection technique to obtain 3D-aware distraction. Furthermore, we incorporate segmentation for pixel-to-segment refinement, enabling more precise identification. Our experiments on benchmark datasets demonstrate that PruNeRF consistently outperforms state-of-the-art methods in robustness against distractors.

In semi-supervised semantic segmentation, the Mean Teacher- and co-training-based approaches are employed to mitigate confirmation bias and coupling problems. However, despite their high performance, these approaches frequently involve complex training pipelines and a substantial computational burden, limiting the scalability and compatibility of these methods. In this paper, we propose a PrevMatch framework that effectively mitigates the aforementioned limitations by maximizing the utilization of the temporal knowledge obtained during the training process. The PrevMatch framework relies on two core strategies: (1) we reconsider the use of temporal knowledge and thus directly utilize previous models obtained during training to generate additional pseudo-label guidance, referred to as previous guidance. (2) we design a highly randomized ensemble strategy to maximize the effectiveness of the previous guidance. Experimental results on four benchmark semantic segmentation datasets confirm that the proposed method consistently outperforms existing methods across various evaluation protocols. In particular, with DeepLabV3+ and ResNet-101 network settings, PrevMatch outperforms the existing state-of-the-art method, Diverse Co-training, by +1.6 mIoU on Pascal VOC with only 92 annotated images, while achieving 2.4 times faster training. Furthermore, the results indicate that PrevMatch induces stable optimization, particularly in benefiting classes that exhibit poor performance. Code is available at https://github.com/wooseok-shin/PrevMatch, Comment: 14 pages, 5 figures, submitted to IEEE TPAMI. This work has been submitted to the IEEE for possible publication

While Large Language Models (LLMs) can serve as agents to simulate human behaviors (i.e., role-playing agents), we emphasize the importance of point-in-time role-playing. This situates characters at specific moments in the narrative progression for three main reasons: (i) enhancing users' narrative immersion, (ii) avoiding spoilers, and (iii) fostering engagement in fandom role-playing. To accurately represent characters at specific time points, agents must avoid character hallucination, where they display knowledge that contradicts their characters' identities and historical timelines. We introduce TimeChara, a new benchmark designed to evaluate point-in-time character hallucination in role-playing LLMs. Comprising 10,895 instances generated through an automated pipeline, this benchmark reveals significant hallucination issues in current state-of-the-art LLMs (e.g., GPT-4o). To counter this challenge, we propose Narrative-Experts, a method that decomposes the reasoning steps and utilizes narrative experts to reduce point-in-time character hallucinations effectively. Still, our findings with TimeChara highlight the ongoing challenges of point-in-time character hallucination, calling for further study., Comment: ACL 2024 Findings. Code and dataset are released at https://ahnjaewoo.github.io/timechara

We present Diffusion Model Patching (DMP), a simple method to boost the performance of pre-trained diffusion models that have already reached convergence, with a negligible increase in parameters. DMP inserts a small, learnable set of prompts into the model's input space while keeping the original model frozen. The effectiveness of DMP is not merely due to the addition of parameters but stems from its dynamic gating mechanism, which selects and combines a subset of learnable prompts at every step of the generative process (e.g., reverse denoising steps). This strategy, which we term "mixture-of-prompts", enables the model to draw on the distinct expertise of each prompt, essentially "patching" the model's functionality at every step with minimal yet specialized parameters. Uniquely, DMP enhances the model by further training on the same dataset on which it was originally trained, even in a scenario where significant improvements are typically not expected due to model convergence. Experiments show that DMP significantly enhances the converged FID of DiT-L/2 on FFHQ 256x256 by 10.38%, achieved with only a 1.43% parameter increase and 50K additional training iterations., Comment: Project page: https://sangminwoo.github.io/DMP/

The multi-plane representation has been highlighted for its fast training and inference across static and dynamic neural radiance fields. This approach constructs relevant features via projection onto learnable grids and interpolating adjacent vertices. However, it has limitations in capturing low-frequency details and tends to overuse parameters for low-frequency features due to its bias toward fine details, despite its multi-resolution concept. This phenomenon leads to instability and inefficiency when training poses are sparse. In this work, we propose a method that synergistically integrates multi-plane representation with a coordinate-based MLP network known for strong bias toward low-frequency signals. The coordinate-based network is responsible for capturing low-frequency details, while the multi-plane representation focuses on capturing fine-grained details. We demonstrate that using residual connections between them seamlessly preserves their own inherent properties. Additionally, the proposed progressive training scheme accelerates the disentanglement of these two features. We demonstrate empirically that our proposed method not only outperforms baseline models for both static and dynamic NeRFs with sparse inputs, but also achieves comparable results with fewer parameters., Comment: ICML2024 ; Project page is accessible at https://mingyukim87.github.io/SynergyNeRF ; Code is available at https://github.com/MingyuKim87/SynergyNeRF

Duality between estimation and control is a foundational concept in Control Theory. Most students learn about the elementary duality -- between observability and controllability -- in their first graduate course in linear systems theory. Therefore, it comes as a surprise that for a more general class of nonlinear stochastic systems (hidden Markov models or HMMs), duality is incomplete. Our objective in writing this article is two-fold: (i) To describe the difficulty in extending duality to HMMs; and (ii) To discuss its recent resolution by the authors. A key message is that the main difficulty in extending duality comes from time reversal in going from estimation to control. The reason for time reversal is explained with the aid of the familiar linear deterministic and linear Gaussian models. The explanation is used to motivate the difference between the linear and the nonlinear models. Once the difference is understood, duality for HMMs is described based on our recent work. The article also includes a comparison and discussion of the different types of duality considered in literature.

Active view selection in 3D scene reconstruction has been widely studied since training on informative views is critical for reconstruction. Recently, Neural Radiance Fields (NeRF) variants have shown promising results in active 3D reconstruction using uncertainty-guided view selection. They utilize uncertainties estimated with neural networks that encode scene geometry and appearance. However, the choice of uncertainty integration methods, either voxel-based or neural rendering, has conventionally depended on the types of scene uncertainty being estimated, whether geometric or appearance-related. In this paper, we introduce Colorized Surface Voxel (CSV)-based view selection, a new next-best view (NBV) selection method exploiting surface voxel-based measurement of uncertainty in scene appearance. CSV encapsulates the uncertainty of estimated scene appearance (e.g., color uncertainty) and estimated geometric information (e.g., surface). Using the geometry information, we interpret the uncertainty of scene appearance 3D-wise during the aggregation of the per-voxel uncertainty. Consequently, the uncertainty from occluded and complex regions is recognized under challenging scenarios with limited input data. Our method outperforms previous works on popular datasets, DTU and Blender, and our new dataset with imbalanced viewpoints, showing that the CSV-based view selection significantly improves performance by up to 30%., Comment: New experiments with newly published dataset are added. The main claims are the same as in the previous version, but the naming and explanations have been changed

In this paper, a backward map is introduced for the purposes of analysis of the nonlinear (stochastic) filter stability. The backward map is important because the filter-stability in the sense of $\chisq$-divergence follows from showing a certain variance decay property for the backward map. To show this property requires additional assumptions on the model properties of the hidden Markov model (HMM). The analysis in this paper is based on introducing a Poincar\'e Inequality (PI) for HMMs with white noise observations. In finite state-space settings, PI is related to both the ergodicity of the Markov process as well as the observability of the HMM. It is shown that the Poincar\'e constant is positive if and only if the HMM is detectable., Comment: Conference proceeding related to arXiv:2305.12850

Human perception involves discerning complex multi-object scenes into time-static object appearance (ie, size, shape, color) and time-varying object motion (ie, location, velocity, acceleration). This innate ability to unconsciously understand the environment is the motivation behind the success of dynamics modeling. Object-centric representations have emerged as a promising tool for dynamics prediction, yet they primarily focus on the objects' appearance, often overlooking other crucial attributes. In this paper, we propose Object-Centric Kinematics (OCK), a framework for dynamics prediction leveraging object-centric representations. Our model utilizes a novel component named object kinematics, which comprises low-level structured states of objects' position, velocity, and acceleration. The object kinematics are obtained via either implicit or explicit approaches, enabling comprehensive spatiotemporal object reasoning, and integrated through various transformer mechanisms, facilitating effective object-centric dynamics modeling. Our model demonstrates superior performance when handling objects and backgrounds in complex scenes characterized by a wide range of object attributes and dynamic movements. Moreover, our model demonstrates generalization capabilities across diverse synthetic environments, highlighting its potential for broad applicability in vision-related tasks., Comment: 15 pages, 6 figures, 4 tables

This paper is divided into two parts. The first part reviews the formulae for f-divergences in the study of continuous-time Markov processes and explores their applications in areas such as stochastic stability, the second law of thermodynamics, and its non-equilibrium extensions. This sets the foundation for the second part, which focuses on f-divergence in the study of hidden Markov processes. In this context, we present analyses of filter stability and stochastic thermodynamics, with the latter being used to illustrate the concept of a Maxwell demon in an over-damped Langevin model with white noise observations. The paper's expository style and unified formalism for both Markov and hidden Markov processes aim to serve as a valuable resource for researchers working across related fields.

We consider the bending of light around a compact astrophysical object with both the electric field and the magnetic field in Einstein-Born-Infeld theory. From the null geodesic of a light ray passing a massive object with electric charge and magnetic dipole, the effective metric was obtained from the light cone condition reflecting the nonlinear electromagnetic effects. We found the asymptotic form of the effective metric up to the first order in gravitational constant $G$ and Born-Infeld parameter $1/\beta^2$ on the equatorial plane. Then we compute the bending angle of light from the geodesic equation. The result includes particular cases where only one type of field is present taking the appropriate limits., Comment: 13 pages, 1 figure, published version in Eur. Phys. J. C

This technical report introduces Pegasus-1, a multimodal language model specialized in video content understanding and interaction through natural language. Pegasus-1 is designed to address the unique challenges posed by video data, such as interpreting spatiotemporal information, to offer nuanced video content comprehension across various lengths. This technical report overviews Pegasus-1's architecture, training strategies, and its performance in benchmarks on video conversation, zero-shot video question answering, and video summarization. We also explore qualitative characteristics of Pegasus-1 , demonstrating its capabilities as well as its limitations, in order to provide readers a balanced view of its current state and its future direction.

We introduce HyperCLOVA X, a family of large language models (LLMs) tailored to the Korean language and culture, along with competitive capabilities in English, math, and coding. HyperCLOVA X was trained on a balanced mix of Korean, English, and code data, followed by instruction-tuning with high-quality human-annotated datasets while abiding by strict safety guidelines reflecting our commitment to responsible AI. The model is evaluated across various benchmarks, including comprehensive reasoning, knowledge, commonsense, factuality, coding, math, chatting, instruction-following, and harmlessness, in both Korean and English. HyperCLOVA X exhibits strong reasoning capabilities in Korean backed by a deep understanding of the language and cultural nuances. Further analysis of the inherent bilingual nature and its extension to multilingualism highlights the model's cross-lingual proficiency and strong generalization ability to untargeted languages, including machine translation between several language pairs and cross-lingual inference tasks. We believe that HyperCLOVA X can provide helpful guidance for regions or countries in developing their sovereign LLMs., Comment: 44 pages; updated authors list and fixed author names

Visual scenes are naturally organized in a hierarchy, where a coarse semantic is recursively comprised of several fine details. Exploring such a visual hierarchy is crucial to recognize the complex relations of visual elements, leading to a comprehensive scene understanding. In this paper, we propose a Visual Hierarchy Mapper (Hi-Mapper), a novel approach for enhancing the structured understanding of the pre-trained Deep Neural Networks (DNNs). Hi-Mapper investigates the hierarchical organization of the visual scene by 1) pre-defining a hierarchy tree through the encapsulation of probability densities; and 2) learning the hierarchical relations in hyperbolic space with a novel hierarchical contrastive loss. The pre-defined hierarchy tree recursively interacts with the visual features of the pre-trained DNNs through hierarchy decomposition and encoding procedures, thereby effectively identifying the visual hierarchy and enhancing the recognition of an entire scene. Extensive experiments demonstrate that Hi-Mapper significantly enhances the representation capability of DNNs, leading to an improved performance on various tasks, including image classification and dense prediction tasks., Comment: This paper is accepted to CVPR 2024. The supplementary material is included. The code is available at \url{https://github.com/kwonjunn01/Hi-Mapper}

Diffusion-based generative models have emerged as powerful tools in the realm of generative modeling. Despite extensive research on denoising across various timesteps and noise levels, a conflict persists regarding the relative difficulties of the denoising tasks. While various studies argue that lower timesteps present more challenging tasks, others contend that higher timesteps are more difficult. To address this conflict, our study undertakes a comprehensive examination of task difficulties, focusing on convergence behavior and changes in relative entropy between consecutive probability distributions across timesteps. Our observational study reveals that denoising at earlier timesteps poses challenges characterized by slower convergence and higher relative entropy, indicating increased task difficulty at these lower timesteps. Building on these observations, we introduce an easy-to-hard learning scheme, drawing from curriculum learning, to enhance the training process of diffusion models. By organizing timesteps or noise levels into clusters and training models with ascending orders of difficulty, we facilitate an order-aware training regime, progressing from easier to harder denoising tasks, thereby deviating from the conventional approach of training diffusion models simultaneously across all timesteps. Our approach leads to improved performance and faster convergence by leveraging benefits of curriculum learning, while maintaining orthogonality with existing improvements in diffusion training techniques. We validate these advantages through comprehensive experiments in image generation tasks, including unconditional, class-conditional, and text-to-image generation.

Diffusion models have achieved remarkable success across a range of generative tasks. Recent efforts to enhance diffusion model architectures have reimagined them as a form of multi-task learning, where each task corresponds to a denoising task at a specific noise level. While these efforts have focused on parameter isolation and task routing, they fall short of capturing detailed inter-task relationships and risk losing semantic information, respectively. In response, we introduce Switch Diffusion Transformer (Switch-DiT), which establishes inter-task relationships between conflicting tasks without compromising semantic information. To achieve this, we employ a sparse mixture-of-experts within each transformer block to utilize semantic information and facilitate handling conflicts in tasks through parameter isolation. Additionally, we propose a diffusion prior loss, encouraging similar tasks to share their denoising paths while isolating conflicting ones. Through these, each transformer block contains a shared expert across all tasks, where the common and task-specific denoising paths enable the diffusion model to construct its beneficial way of synergizing denoising tasks. Extensive experiments validate the effectiveness of our approach in improving both image quality and convergence rate, and further analysis demonstrates that Switch-DiT constructs tailored denoising paths across various generation scenarios., Comment: Project Page: https://byeongjun-park.github.io/Switch-DiT/