Author name cluster

Naiqi Li

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

17 papers

2 author rows

AAAI Conference 2025 Conference Paper

CALF: Aligning LLMs for Time Series Forecasting via Cross-modal Fine-Tuning

Peiyuan Liu
Hang Guo
Tao Dai
Naiqi Li
Jigang Bao
Xudong Ren
Yong Jiang
Shu-Tao Xia

Deep learning (e.g., Transformer) has been widely and successfully used in multivariate time series forecasting (MTSF). Unlike existing methods that focus on training models from a single modal of time series input, large language models (LLMs) based MTSF methods with cross-modal text and time series input have recently shown great superiority, especially with limited temporal data. However, current LLM-based MTSF methods usually focus on adapting and fine-tuning LLMs, while neglecting the distribution discrepancy between textual and temporal input tokens, thus leading to sub-optimal performance. To address this issue, we propose a novel Cross-Modal LLM Fine-Tuning (CALF) framework for MTSF by reducing the distribution discrepancy between textual and temporal data, which mainly consists of the temporal target branch with temporal input and the textual source branch with aligned textual input. To reduce the distribution discrepancy, we develop the cross-modal match module to first align cross-modal input distributions. Additionally, to minimize the modality distribution gap in both feature and output spaces, feature regularization loss is developed to align the intermediate features between the two branches for better weight updates, while output consistency loss is introduced to allow the output representations of both branches to correspond effectively. Thanks to the modality alignment, CALF establishes state-of-the-art performance for both long-term and short-term forecasting tasks with low computational complexity, and exhibits favorable few-shot and zero-shot abilities similar to that in LLMs.

PDF Details DOI

AAAI Conference 2025 Conference Paper

Diffusion Prior Interpolation for Flexibility Real-World Face Super-Resolution

Jiarui Yang
Tao Dai
Yufei Zhu
Naiqi Li
Jinmin Li
Shu-Tao Xia

Diffusion models represent the state-of-the-art in generative modeling. Due to their high training costs, many works leverage pre-trained diffusion models' powerful representations for downstream tasks, such as face super-resolution (FSR), through fine-tuning or prior-based methods. However, relying solely on priors without supervised training makes it challenging to meet the pixel-level accuracy requirements of discrimination task. Although prior-based methods can achieve high fidelity and high-quality results, ensuring consistency remains a significant challenge. In this paper, we propose a masking strategy with strong and weak constraints and iterative refinement for real-world FSR, termed Diffusion Prior Interpolation (DPI). We introduce conditions and constraints on consistency by masking different sampling stages based on the structural characteristics of the face. Furthermore, we propose a condition Corrector (CRT) to establish a reciprocal posterior sampling process. DPI can balance consistency and diversity and can be seamlessly integrated into pre-trained models. In extensive experiments conducted on synthetic and real datasets, along with consistency validation in face recognition, DPI demonstrates superiority over SOTA FSR methods.

PDF Details DOI

IJCAI Conference 2025 Conference Paper

Efficient Differentiable Approximation of Generalized Low-rank Regularization

Naiqi Li
Yuqiu Xie
Peiyuan Liu
Tao Dai
Yong Jiang
Shu-Tao Xia

Low-rank regularization (LRR) has been widely applied in various machine learning tasks, but the associated optimization is challenging. Directly optimizing the rank function under constraints is NP-hard in general. To overcome this difficulty, various relaxations of the rank function were studied. However, optimization of these relaxed LRRs typically depends on singular value decomposition, which is a time-consuming and nondifferentiable operator that cannot be optimized with gradient-based techniques. To address these challenges, in this paper we propose an efficient differentiable approximation of the generalized LRR. The considered LRR form subsumes many popular choices like the nuclear norm, the Schatten-p norm, and various nonconvex relaxations. Our method enables LRR terms to be appended to loss functions in a plug-and-play fashion, and the GPU-friendly operations enable efficient and convenient implementation. Furthermore, convergence analysis is presented, which rigorously shows that both the bias and the variance of our rank estimator rapidly reduce with increased sample size and iteration steps. In the experimental study, the proposed method is applied to various tasks, which demonstrates its versatility and efficiency. Code is available at https: //github. com/naiqili/EDLRR.

PDF Details DOI

ICRA Conference 2025 Conference Paper

Expert-Enhanced Masked Point Modeling for Point Cloud Self-Supervised Learning

Yujun Liu
Yaohua Zha
Naiqi Li
Tao Dai 0001
Bin Chen 0011
Shu-Tao Xia

Recently, learning-based point cloud analysis has played a crucial role in robotic perception. Masked Point Modeling (MPM), owing to its powerful representational capabilities, has become the mainstream point cloud self-supervised learning method. However, existing MPM-based methods often suffer from the problem of negative transfer, due to the disparity in semantic distribution between upstream data and downstream data. To address this issue, we propose an expert enhancement strategy for existing MPM-based methods. Specifically, we insert a Sparse Mixture of Experts (SMoE) layer after each block of the backbone network, which utilizes a multi-branch expert architecture with routers that allocate data of different semantics to the appropriate experts for analysis. During the pre-training phase, our expert-enhanced model not only learns universal 3D representations for the backbone network but also acquires powerful semantic routing capabilities for all expert layers. In the fine-tuning phase, we freeze all backbones and conduct end-to-end fine-tuning solely on our expert layers to adaptively select multiple experts most relevant to the semantics of each downstream data for analysis. Extensive downstream experiments demonstrate the superiority of our method, especially outperforming baseline (Point-MAE) by 5. 16%, 5. 86%, and 4. 62% in three variants of ScanObjectNN while utilizing only 12% of its trainable parameters. Our code is released at https://github.com/chenchen1104/point_e2mae.

Details

ICML Conference 2025 Conference Paper

TimeBridge: Non-Stationarity Matters for Long-term Time Series Forecasting

Peiyuan Liu
Beiliang Wu
Yifan Hu 0006
Naiqi Li
Tao Dai 0001
Jigang Bao
Shu-Tao Xia

Non-stationarity poses significant challenges for multivariate time series forecasting due to the inherent short-term fluctuations and long-term trends that can lead to spurious regressions or obscure essential long-term relationships. Most existing methods either eliminate or retain non-stationarity without adequately addressing its distinct impacts on short-term and long-term modeling. Eliminating non-stationarity is essential for avoiding spurious regressions and capturing local dependencies in short-term modeling, while preserving it is crucial for revealing long-term cointegration across variates. In this paper, we propose TimeBridge, a novel framework designed to bridge the gap between non-stationarity and dependency modeling in long-term time series forecasting. By segmenting input series into smaller patches, TimeBridge applies Integrated Attention to mitigate short-term non-stationarity and capture stable dependencies within each variate, while Cointegrated Attention preserves non-stationarity to model long-term cointegration across variates. Extensive experiments show that TimeBridge consistently achieves state-of-the-art performance in both short-term and long-term forecasting. Additionally, TimeBridge demonstrates exceptional performance in financial forecasting on the CSI 500 and S&P 500 indices, further validating its robustness and effectiveness. Code is available at https: //github. com/Hank0626/TimeBridge.

Details

ICML Conference 2025 Conference Paper

TimeFilter: Patch-Specific Spatial-Temporal Graph Filtration for Time Series Forecasting

Yifan Hu 0006
Guibin Zhang
Peiyuan Liu
Disen Lan
Naiqi Li
Dawei Cheng
Tao Dai 0001
Shu-Tao Xia

Time series forecasting methods generally fall into two main categories: Channel Independent (CI) and Channel Dependent (CD) strategies. While CI overlooks important covariate relationships, CD captures all dependencies without distinction, introducing noise and reducing generalization. Recent advances in Channel Clustering (CC) aim to refine dependency modeling by grouping channels with similar characteristics and applying tailored modeling techniques. However, coarse-grained clustering struggles to capture complex, time-varying interactions effectively. To address these challenges, we propose TimeFilter, a GNN-based framework for adaptive and fine-grained dependency modeling. After constructing the graph from the input sequence, TimeFilter refines the learned spatial-temporal dependencies by filtering out irrelevant correlations while preserving the most critical ones in a patch-specific manner. Extensive experiments on 13 real-world datasets from diverse application domains demonstrate the state-of-the-art performance of TimeFilter. The code is available at https: //github. com/TROUBADOUR000/TimeFilter.

Details

NeurIPS Conference 2024 Conference Paper

DDN: Dual-domain Dynamic Normalization for Non-stationary Time Series Forecasting

Tao Dai
Beiliang Wu
Peiyuan Liu
Naiqi Li
Xue Yuerong
Shu-Tao Xia
Zexuan Zhu

Deep neural networks (DNNs) have recently achieved remarkable advancements in time series forecasting (TSF) due to their powerful ability of sequence dependence modeling. To date, existing DNN-based TSF methods still suffer from unreliable predictions for real-world data due to its non-stationarity characteristics, i. e. , data distribution varies quickly over time. To mitigate this issue, several normalization methods (e. g. , SAN) have recently been specifically designed by normalization in a fixed period/window in the time domain. However, these methods still struggle to capture distribution variations, due to the complex time patterns of time series in the time domain. Based on the fact that wavelet transform can decompose time series into a linear combination of different frequencies, which exhibits distribution variations with time-varying periods, we propose a novel Dual-domain Dynamic Normalization (DDN) to dynamically capture distribution variations in both time and frequency domains. Specifically, our DDN tries to eliminate the non-stationarity of time series via both frequency and time domain normalization in a sliding window way. Besides, our DDN can serve as a plug-in-play module, and thus can be easily incorporated into other forecasting models. Extensive experiments on public benchmark datasets under different forecasting models demonstrate the superiority of our DDN over other normalization methods. Code will be made available following the review process.

PDF Details DOI

IJCAI Conference 2024 Conference Paper

GladCoder: Stylized QR Code Generation with Grayscale-Aware Denoising Process

Yuqiu Xie
Bolin Jiang
Jiawei Li
Naiqi Li
Bin Chen
Tao Dai
Yuang Peng
Shu-Tao Xia

Traditional QR codes consist of a grid of black-and-white square modules, which lack aesthetic appeal and meaning for human perception. This has motivated recent research to beautify the visual appearance of QR codes. However, there exists a trade-off between the visual quality and scanning-robustness of the image, causing outputs of previous works are simple and of low quality to ensure scanning-robustness. In this paper, we introduce a novel approach GladCoder to generate stylized QR codes that are personalized, natural, and text-driven. Its pipeline includes a Depth-guided Aesthetic QR code Generator (DAG) to improve quality of image foreground, and a GrayscaLe-Aware Denoising (GLAD) process to enhance scanning-robustness. The overall pipeline is based on diffusion models, which allow users to create stylized QR images from a textual prompt to describe the image and a textual input to be encoded. Experiments demonstrate that our method can generate stylized QR code with appealing perception details, while maintaining robust scanning reliability under real world applications.

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

LCM: Locally Constrained Compact Point Cloud Model for Masked Point Modeling

Yaohua Zha
Naiqi Li
Yanzi Wang
Tao Dai
Hang Guo
Bin Chen
Zhi Wang
Zhihao Ouyang

The pre-trained point cloud model based on Masked Point Modeling (MPM) has exhibited substantial improvements across various tasks. However, these models heavily rely on the Transformer, leading to quadratic complexity and limited decoder, hindering their practice application. To address this limitation, we first conduct a comprehensive analysis of existing Transformer-based MPM, emphasizing the idea that redundancy reduction is crucial for point cloud analysis. To this end, we propose a Locally constrained Compact point cloud Model (LCM) consisting of a locally constrained compact encoder and a locally constrained Mamba-based decoder. Our encoder replaces self-attention with our local aggregation layers to achieve an elegant balance between performance and efficiency. Considering the varying information density between masked and unmasked patches in the decoder inputs of MPM, we introduce a locally constrained Mamba-based decoder. This decoder ensures linear complexity while maximizing the perception of point cloud geometry information from unmasked patches with higher information density. Extensive experimental results show that our compact model significantly surpasses existing Transformer-based models in both performance and efficiency, especially our LCM-based Point-MAE model, compared to the Transformer-based model, achieved an improvement of 1. 84%, 0. 67%, and 0. 60% in performance on the three variants of ScanObjectNN while reducing parameters by 88% and computation by 73%. The code is available at https: //github. com/zyh16143998882/LCM.

PDF Details DOI

ICLR Conference 2024 Conference Paper

Periodicity Decoupling Framework for Long-term Series Forecasting

Tao Dai 0001
Beiliang Wu
Peiyuan Liu
Naiqi Li
Jigang Bao
Yong Jiang 0001
Shu-Tao Xia

Convolutional neural network (CNN)-based and Transformer-based methods have recently made significant strides in time series forecasting, which excel at modeling local temporal variations or capturing long-term dependencies. However, real-world time series usually contain intricate temporal patterns, thus making it challenging for existing methods that mainly focus on temporal variations modeling from the 1D time series directly. Based on the intrinsic periodicity of time series, we propose a novel Periodicity Decoupling Framework (PDF) to capture 2D temporal variations of decoupled series for long-term series forecasting. Our PDF mainly consists of three components: multi-periodic decoupling block (MDB), dual variations modeling block (DVMB), and variations aggregation block (VAB). Unlike the previous methods that model 1D temporal variations, our PDF mainly models 2D temporal variations, decoupled from 1D time series by MDB. After that, DVMB attempts to further capture short-term and long-term variations, followed by VAB to make final predictions. Extensive experimental results across seven real-world long-term time series datasets demonstrate the superiority of our method over other state-of-the-art methods, in terms of both forecasting performance and computational efficiency. Code is available at https://github.com/Hank0626/PDF.

Details

AAAI Conference 2024 Conference Paper

Procedural Level Generation with Diffusion Models from a Single Example

Shiqi Dai
Xuanyu Zhu
Naiqi Li
Tao Dai
Zhi Wang

Level generation is a central focus of Procedural Content Generation (PCG), yet deep learning-based approaches are limited by scarce training data, i.e., human-designed levels. Despite being a dominant framework, Generative Adversarial Networks (GANs) exhibit a substantial quality gap between generated and human-authored levels, alongside rising training costs, particularly with increasing token complexity. In this paper, we introduce a diffusion-based generative model that learns from just one example. Our approach involves two core components: 1) an efficient yet expressive level representation, and 2) a latent denoising network with constrained receptive fields. To start with, our method utilizes token semantic labels, similar to word embeddings, to provide dense representations. This strategy not only surpasses one-hot encoding in representing larger game levels but also improves stability and accelerates convergence in latent diffusion. In addition, we adapt the denoising network architecture to confine the receptive field to localized patches of the data, aiming to facilitate single-example learning. Extensive experiments demonstrate that our model is capable of generating stylistically congruent samples of arbitrary sizes compared to manually designed levels. It suits a wide range of level structures with fewer artifacts than GAN-based approaches. The source code is available at https://github.com/shiqi-dai/diffusioncraft.

PDF Details DOI

UAI Conference 2022 Conference Paper

Deep Dirichlet process mixture models

Naiqi Li
Wenjie Li 0008
Yong Jiang 0001
Shu-Tao Xia

In this paper we propose the deep Dirichlet process mixture (DDPM) model, which is an unsupervised method that simultaneously performs clustering and feature learning. The traditional Dirichlet process mixture model can infer the number of mixture components, but its flexibility is restricted since the clustering is performed in the raw feature space. Our method alleviates this limitation by using the flow-based deep neural network to learn more expressive features. DDPM unifies Dirichlet processes and the flow-based model with Monte Carlo expectation-maximization, and uses Gibbs sampling to sample from the posterior. This combination allows our method to exploit the mutually beneficial relation between clustering and feature learning. The effectiveness of DDPM is demonstrated by thorough experiments in various synthetic and real-world datasets.

Details

NeurIPS Conference 2020 Conference Paper

Stochastic Deep Gaussian Processes over Graphs

Naiqi Li
Wenjie Li
Jifeng Sun
Yinghua Gao
Yong Jiang
Shu-Tao Xia

In this paper we propose Stochastic Deep Gaussian Processes over Graphs (DGPG), which are deep structure models that learn the mappings between input and output signals in graph domains. The approximate posterior distributions of the latent variables are derived with variational inference, and the evidence lower bound is evaluated and optimized by the proposed recursive sampling scheme. The Bayesian non-parametric natural of our model allows it to resist overfitting, while the expressive deep structure grants it the potential to learn complex relations. Extensive experiments demonstrate that our method achieves superior performances in both small size ( 35, 000) datasets. We show that DGPG outperforms another Gaussian-based approach, and is competitive to a state-of-the-art method in the challenging task of traffic flow prediction. Our model is also capable of capturing uncertainties in a mathematical principled way and automatically discovering which vertices and features are relevant to the prediction.

PDF Details

ECAI Conference 2016 Conference Paper

Automatic Verification of Golog Programs via Predicate Abstraction

Peiming Mo
Naiqi Li
Yongmei Liu 0001

Golog is a logic programming language for high-level agent control. In a recent paper, we proposed a sound but incomplete method for automatic verification of partial correctness of Golog programs where we give a number of heuristic methods to strengthen given formulas in order to discover loop invariants. However, our method does not work on arithmetic domains. On the other hand, the method of predicate abstraction is widely used in the software engineering community for model checking and partial correctness verification of programs. Intuitively, the predicate abstraction task is to find a formula consisting of a given set of predicates to approximate a given first-order formula. In this paper, we propose a method for automatic verification of partial correctness of Golog programs which use predicate abstraction as a uniform method to strengthen given formulas. We implement a system based on the proposed method, conduct experiments on arithmetical domains and examples from the paper by Li and Liu. Also, we apply our method to the verification of winning strategies for combinatorial games.

Details

IJCAI Conference 2015 Conference Paper

Automatic Verification of Partial Correctness of Golog Programs

Naiqi Li
Yongmei Liu

When Golog programs are used to control agents’ behaviour in a high-level manner, their partial correctness naturally becomes an important concern. In this paper we propose a sound but incomplete method for automatic verification of partial correctness of Golog programs. We introduce the notion of extended regression, which reduces partial correctness of Golog programs to first-order entailment problems. During the process loop invariants are automatically discovered by heuristic methods. We propose progression of small models wrt Golog programs, which are used to filter out too strong heuristic candidates. In this way we combine the methods of static and dynamic analysis from the software engineering community. Furthermore, our method can also be adapted to verify state constraints. Experiments show that our method can not only handle sequential and nested loops uniformly in a reasonable among of time, but also be used to discover succinct and comprehensible loop invariants and state constraints.

PDF Details

IJCAI Conference 2013 Conference Paper

Reasoning about State Constraints in the Situation Calculus

Naiqi Li
Yi Fan
Yongmei Liu

In dynamic systems, state constraints are formulas that hold in every reachable state. It has been shown that state constraints can be used to greatly reduce the planning search space. They are also useful in program verification. In this paper, we propose a sound but incomplete method for automatic verification and discovery of ∀∗ ∃∗ state constraints for a class of action theories that include many planning benchmarks. Our method is formulated in the situation calculus, theoretically based on Skolemization and Herbrand Theorem, and implemented with SAT solvers. Basically, we verify a state constraint by strengthening it in a novel and smart way so that it becomes a state invariant. We experimented with the blocks world, logistics and satellite domains, and the results showed that, almost all known state constraints can be verified in a reasonable amount of time, and meanwhile succinct and intuitive related state constraints are discovered.

PDF Details DOI

AAAI Conference 2012 Conference Paper

A First-Order Interpreter for Knowledge-Based Golog with Sensing based on Exact Progression and Limited Reasoning

Yi Fan
Minghui Cai
Naiqi Li
Yongmei Liu

While founded on the situation calculus, current implementations of Golog are mainly based on the closedworld assumption or its dynamic versions or the domain closure assumption. Also, they are almost exclusively based on regression. In this paper, we propose a first-order interpreter for knowledge-based Golog with sensing based on exact progression and limited reasoning. We assume infinitely many unique names and handle first-order disjunctive information in the form of the so-called proper+ KBs. Our implementation is based on the progression and limited reasoning algorithms for proper+ KBs proposed by Liu, Lakemeyer and Levesque. To improve efficiency, we implement the two algorithms by grounding via a trick based on the unique name assumption. The interpreter is online but the programmer can use two operators to specify offline execution for parts of programs. The search operator returns a conditional plan, while the planning operator is used when local closed-world information is available and calls a modern planner to generate a sequence of actions.

PDF Details