arXiv subjects

Dan Zhang

Publications and source records attributed to Dan Zhang.

At least 145 records · Page 8Linked to original sources

arXiv2024

Unleashing the Power of Self-Supervised Image Denoising: A Comprehensive Review

The advent of deep learning has brought a revolutionary transformation to image denoising techniques. However, the persistent challenge of acquiring noise-clean pairs for supervised methods in real-world scenarios remains formidable, necessitating the exploration of more practical self-supervised image denoising. This paper focuses on self-supervised image denoising methods that offer effective solutions to address this challenge. Our comprehensive review thoroughly analyzes the latest advancements in self-supervised image denoising approaches, categorizing them into three distinct classes: General methods, Blind Spot Network (BSN)-based methods, and Transformer-based methods. For each class, we provide a concise theoretical analysis along with their practical applications. To assess the effectiveness of these methods, we present both quantitative and qualitative experimental results on various datasets, utilizing classical algorithms as benchmarks. Additionally, we critically discuss the current limitations of these methods and propose promising directions for future research. By offering a detailed overview of recent developments in self-supervised image denoising, this review serves as an invaluable resource for researchers and practitioners in the field, facilitating a deeper understanding of this emerging domain and inspiring further advancements.

eess.IV↗

arXiv2024

Controlling Text-to-Image Diffusion by Orthogonal Finetuning

Large text-to-image diffusion models have impressive capabilities in generating photorealistic images from text prompts. How to effectively guide or control these powerful models to perform different downstream tasks becomes an important open problem. To tackle this challenge, we introduce a principled finetuning method -- Orthogonal Finetuning (OFT), for adapting text-to-image diffusion models to downstream tasks. Unlike existing methods, OFT can provably preserve hyperspherical energy which characterizes the pairwise neuron relationship on the unit hypersphere. We find that this property is crucial for preserving the semantic generation ability of text-to-image diffusion models. To improve finetuning stability, we further propose Constrained Orthogonal Finetuning (COFT) which imposes an additional radius constraint to the hypersphere. Specifically, we consider two important finetuning text-to-image tasks: subject-driven generation where the goal is to generate subject-specific images given a few images of a subject and a text prompt, and controllable generation where the goal is to enable the model to take in additional control signals. We empirically show that our OFT framework outperforms existing methods in generation quality and convergence speed.

cs.CV↗

arXiv2024

Language-Specific Representation of Emotion-Concept Knowledge Causally Supports Emotion Inference

Humans no doubt use language to communicate about their emotional experiences, but does language in turn help humans understand emotions, or is language just a vehicle of communication? This study used a form of artificial intelligence (AI) known as large language models (LLMs) to assess whether language-based representations of emotion causally contribute to the AI's ability to generate inferences about the emotional meaning of novel situations. Fourteen attributes of human emotion concept representation were found to be represented by the LLM's distinct artificial neuron populations. By manipulating these attribute-related neurons, we in turn demonstrated the role of emotion concept knowledge in generative emotion inference. The attribute-specific performance deterioration was related to the importance of different attributes in human mental space. Our findings provide a proof-in-concept that even a LLM can learn about emotions in the absence of sensory-motor representations and highlight the contribution of language-derived emotion-concept knowledge for emotion inference.

cs.AI↗

arXiv2024

Detecting Neutrinos from Supernova Bursts in PandaX-4T

Neutrinos from core-collapse supernovae are essential for the understanding of neutrino physics and stellar evolution. The dual-phase xenon dark matter detectors can provide a way to track explosions of galactic supernovae by detecting neutrinos through coherent elastic neutrino-nucleus scatterings. In this study, a variation of progenitor masses as well as explosion models are assumed to predict the neutrino fluxes and spectra, which result in the number of expected neutrino events ranging from 6.6 to 13.7 at a distance of 10 kpc over a 10-second duration with negligible backgrounds at PandaX-4T. Two specialized triggering alarms for monitoring supernova burst neutrinos are built. The efficiency of detecting supernova explosions at various distances in the Milky Way is estimated. These alarms will be implemented in the real-time supernova monitoring system at PandaX-4T in the near future, providing the astronomical communities with supernova early warnings.

hep-ex↗

arXiv2024

Quasinormal modes of a regular black hole with sub-Planckian curvature

This paper explores the properties of the quasinormal modes (QNMs) of a regular black hole(BH) characterized by a Minkowski core and sub-Planckian curvature. When focusing on a special case, this regular BH exhibits identical large-scale behavior with the Hayward BH and some loop quantum gravity corrected (LQG-corrected) BH. A notable characteristic of the QNMs in this regular BH is the pronounced outburst of overtones when compared to the Schwarzschild BH (SS-BH). This outburst can be attributed to the deviation from the SS-BH in the near-horizon geometry region due to the quantum gravity effect. Furthermore, we compare the QNM properties of the regular BH with those of the Hayward BH and the LQG-corrected BH. A similar phenomenon of overtone outburst is observed in the modes of the overtone. As a conclusion, the QNMs may be a powerful tool for detecting the quantum gravity effect and distinguishing different BH models.

gr-qc↗

arXiv2024

MEGAnno+: A Human-LLM Collaborative Annotation System

Large language models (LLMs) can label data faster and cheaper than humans for various NLP tasks. Despite their prowess, LLMs may fall short in understanding of complex, sociocultural, or domain-specific context, potentially leading to incorrect annotations. Therefore, we advocate a collaborative approach where humans and LLMs work together to produce reliable and high-quality labels. We present MEGAnno+, a human-LLM collaborative annotation system that offers effective LLM agent and annotation management, convenient and robust LLM annotation, and exploratory verification of LLM labels by humans.

cs.CL↗

arXiv2024

RecDCL: Dual Contrastive Learning for Recommendation

Self-supervised learning (SSL) has recently achieved great success in mining the user-item interactions for collaborative filtering. As a major paradigm, contrastive learning (CL) based SSL helps address data sparsity in Web platforms by contrasting the embeddings between raw and augmented data. However, existing CL-based methods mostly focus on contrasting in a batch-wise way, failing to exploit potential regularity in the feature dimension. This leads to redundant solutions during the representation learning of users and items. In this work, we investigate how to employ both batch-wise CL (BCL) and feature-wise CL (FCL) for recommendation. We theoretically analyze the relation between BCL and FCL, and find that combining BCL and FCL helps eliminate redundant solutions but never misses an optimal solution. We propose a dual contrastive learning recommendation framework -- RecDCL. In RecDCL, the FCL objective is designed to eliminate redundant solutions on user-item positive pairs and to optimize the uniform distributions within users and items using a polynomial kernel for driving the representations to be orthogonal; The BCL objective is utilized to generate contrastive embeddings on output vectors for enhancing the robustness of the representations. Extensive experiments on four widely-used benchmarks and one industry dataset demonstrate that RecDCL can consistently outperform the state-of-the-art GNNs-based and SSL-based models (with an improvement of up to 5.65\% in terms of Recall@20). The source code is publicly available (https://github.com/THUDM/RecDCL).

cs.IR↗

arXiv2024

Knowledge Acquisition and Integration with Expert-in-the-loop

Constructing and serving knowledge graphs (KGs) is an iterative and human-centered process involving on-demand programming and analysis. In this paper, we present Kyurem, a programmable and interactive widget library that facilitates human-in-the-loop knowledge acquisition and integration to enable continuous curation a knowledge graph (KG). Kyurem provides a seamless environment within computational notebooks where data scientists explore a KG to identify opportunities for acquiring new knowledge and verify recommendations provided by AI agents for integrating the acquired knowledge in the KG. We refined Kyurem through participatory design and conducted case studies in a real-world setting for evaluation. The case-studies show that introduction of Kyurem within an existing HR knowledge graph construction and serving platform improved the user experience of the experts and helped eradicate inefficiencies related to knowledge acquisition and integration tasks

cs.HC↗

arXiv2024

Adversarial Supervision Makes Layout-to-Image Diffusion Models Thrive

Despite the recent advances in large-scale diffusion models, little progress has been made on the layout-to-image (L2I) synthesis task. Current L2I models either suffer from poor editability via text or weak alignment between the generated image and the input layout. This limits their usability in practice. To mitigate this, we propose to integrate adversarial supervision into the conventional training pipeline of L2I diffusion models (ALDM). Specifically, we employ a segmentation-based discriminator which provides explicit feedback to the diffusion generator on the pixel-level alignment between the denoised image and the input layout. To encourage consistent adherence to the input layout over the sampling steps, we further introduce the multistep unrolling strategy. Instead of looking at a single timestep, we unroll a few steps recursively to imitate the inference process, and ask the discriminator to assess the alignment of denoised images with the layout over a certain time window. Our experiments show that ALDM enables layout faithfulness of the generated images, while allowing broad editability via text prompts. Moreover, we showcase its usefulness for practical applications: by synthesizing target distribution samples via text control, we improve domain generalization of semantic segmentation models by a large margin (~12 mIoU points).

cs.CV↗

arXiv2024

Quasinormal modes of quantum-corrected black holes

In this paper, we investigate the quasinormal mode (QNM) spectra for scalar perturbation over a quantum-corrected black hole (BH). The fundamental modes of this quantum-corrected BH exhibit two key properties. Firstly, there is a non-monotonic behavior concerning the quantum-corrected parameter for zero multipole number. Secondly, the quantum gravity effects result in slower decay modes. For higher overtones, a significant deviation becomes evident between the quasinormal frequencies (QNFs) of the quantum-corrected and Schwarzschild BHs. The intervention of quantum gravity corrections induces a significant outburst of overtones. This outburst of these overtones can be attributed to the distinctions near the event horizons between the Schwarzschild and quantum-corrected BHs. Therefore, overtones can serve as a means to probe physical phenomena or disparities in the vicinity of the event horizon.

gr-qc↗

arXiv2023

Search for Dark-Matter-Nucleon Interactions with a Dark Mediator in PandaX-4T

We report results of a search for dark-matter-nucleon interactions via a dark mediator using optimized low-energy data from the PandaX-4T liquid xenon experiment. With the ionization-signal-only data and utilizing the Migdal effect, we set the most stringent limits on the cross section for dark matter masses ranging from 30~$\rm{MeV/c^2}$ to 2~$\rm{GeV/c^2}$. Under the assumption that the dark mediator is a dark photon that decays into scalar dark matter pairs in the early Universe, we rule out significant parameter space of such thermal relic dark-matter model.

hep-ex↗

arXiv2023

Generating Novel Scene Compositions from Single Images and Videos

Given a large dataset for training, generative adversarial networks (GANs) can achieve remarkable performance for the image synthesis task. However, training GANs in extremely low data regimes remains a challenge, as overfitting often occurs, leading to memorization or training divergence. In this work, we introduce SIV-GAN, an unconditional generative model that can generate new scene compositions from a single training image or a single video clip. We propose a two-branch discriminator architecture, with content and layout branches designed to judge internal content and scene layout realism separately from each other. This discriminator design enables synthesis of visually plausible, novel compositions of a scene, with varying content and layout, while preserving the context of the original sample. Compared to previous single image GANs, our model generates more diverse, higher quality images, while not being restricted to a single image setting. We further introduce a new challenging task of learning from a few frames of a single video. In this training setup the training images are highly similar to each other, which makes it difficult for prior GAN models to achieve a synthesis of both high quality and diversity.

cs.CV↗

arXiv2023

Beyond Top-Class Agreement: Using Divergences to Forecast Performance under Distribution Shift

Knowing if a model will generalize to data 'in the wild' is crucial for safe deployment. To this end, we study model disagreement notions that consider the full predictive distribution - specifically disagreement based on Hellinger distance, Jensen-Shannon and Kullback-Leibler divergence. We find that divergence-based scores provide better test error estimates and detection rates on out-of-distribution data compared to their top-1 counterparts. Experiments involve standard vision and foundation models.

cs.LG↗

arXiv2023

Peculiar properties in quasi-normal spectra from loop quantum gravity effect

We investigate the quasi-normal mode (QNM) spectra for scalar and electromagnetic fields over a covairant loop quantum gravity black hole (LQG-BH). For the fundamental modes, the LQG effect reduces the oscillations in the scalar field, however it induces stronger oscillations in the electromagnetic field, comparing to the classical case. Under the scalar field perturbation, the system enjoys faster decaying modes with more oscillations than the electromagnetic field. Some peculiar phenomena emerge in the QNM spectra with higher overtones. A notable feature is the substantial divergence observed in the first several overtones from their Schwarzschild counterparts, with this discrepancy magnifying as the overtone number increases. Another remarkable phenomenon in higher overtones is that the QNFs of the scalar field with $l=0$ exhibit an oscillatory behavior as the quantum parameter $r_0$ increases significantly. These oscillations intensify with the rising overtone number. We hypothesize that this oscillatory pattern may be associated with the extremal effect.

gr-qc↗

arXiv2023

Learning Sample Difficulty from Pre-trained Models for Reliable Prediction

Large-scale pre-trained models have achieved remarkable success in many applications, but how to leverage them to improve the prediction reliability of downstream models is undesirably under-explored. Moreover, modern neural networks have been found to be poorly calibrated and make overconfident predictions regardless of inherent sample difficulty and data uncertainty. To address this issue, we propose to utilize large-scale pre-trained models to guide downstream model training with sample difficulty-aware entropy regularization. Pre-trained models that have been exposed to large-scale datasets and do not overfit the downstream training classes enable us to measure each training sample's difficulty via feature-space Gaussian modeling and relative Mahalanobis distance computation. Importantly, by adaptively penalizing overconfident prediction based on the sample difficulty, we simultaneously improve accuracy and uncertainty calibration across challenging benchmarks (e.g., +0.55% ACC and -3.7% ECE on ImageNet1k using ResNet34), consistently surpassing competitive baselines for reliable prediction. The improved uncertainty estimate further improves selective classification (abstaining from erroneous predictions) and out-of-distribution detection.

cs.LG↗

arXiv2023

Towards Anytime Classification in Early-Exit Architectures by Enforcing Conditional Monotonicity

Modern predictive models are often deployed to environments in which computational budgets are dynamic. Anytime algorithms are well-suited to such environments as, at any point during computation, they can output a prediction whose quality is a function of computation time. Early-exit neural networks have garnered attention in the context of anytime computation due to their capability to provide intermediate predictions at various stages throughout the network. However, we demonstrate that current early-exit networks are not directly applicable to anytime settings, as the quality of predictions for individual data points is not guaranteed to improve with longer computation. To address this shortcoming, we propose an elegant post-hoc modification, based on the Product-of-Experts, that encourages an early-exit network to become gradually confident. This gives our deep models the property of conditional monotonicity in the prediction quality -- an essential stepping stone towards truly anytime predictive modeling using early-exit architectures. Our empirical results on standard image-classification tasks demonstrate that such behaviors can be achieved while preserving competitive accuracy on average.

cs.LG↗

arXiv2023

DocPrompt: Large-scale continue pretrain for zero-shot and few-shot document question answering

In this paper, we propose Docprompt for document question answering tasks with powerful zero-shot and few-shot performance. We proposed a novel weakly supervised data generation method, a novel multl-stage training method and a novel understanding model \& generation model ensemble method. We achieved state-of-the-art performance on 4 document question answering tasks. This method greatly improves the delivery efficiency and model performance of document question answering customer projects, reducing annotation costs and labor costs. Our demo can be found at https://huggingface.co/spaces/PaddlePaddle/ERNIE-Layout.

cs.CL↗

arXiv2023

TriangleNet: Edge Prior Augmented Network for Semantic Segmentation through Cross-Task Consistency

This paper addresses the task of semantic segmentation in computer vision, aiming to achieve precise pixel-wise classification. We investigate the joint training of models for semantic edge detection and semantic segmentation, which has shown promise. However, implicit cross-task consistency learning in multi-task networks is limited. To address this, we propose a novel "decoupled cross-task consistency loss" that explicitly enhances cross-task consistency. Our semantic segmentation network, TriangleNet, achieves a substantial 2.88\% improvement over the Baseline in mean Intersection over Union (mIoU) on the Cityscapes test set. Notably, TriangleNet operates at 77.4\% mIoU/46.2 FPS on Cityscapes, showcasing real-time inference capabilities at full resolution. With multi-scale inference, performance is further enhanced to 77.8\%. Furthermore, TriangleNet consistently outperforms the Baseline on the FloodNet dataset, demonstrating its robust generalization capabilities. The proposed method underscores the significance of multi-task learning and explicit cross-task consistency enhancement for advancing semantic segmentation and highlights the potential of multitasking in real-time semantic segmentation.

cs.CV↗