Search

We propose a compressive yet effective mesh representation, Blocked and Patchified Tokenization (BPT), facilitating the generation of meshes exceeding 8k faces. BPT compresses mesh sequences by employing block-wise indexing and patch aggregation, reducing their length by approximately 75\% compared to the original sequences. This compression milestone unlocks the potential to utilize mesh data with significantly more faces, thereby enhancing detail richness and improving generation robustness. Empowered with the BPT, we have built a foundation mesh generative model training on scaled mesh data to support flexible control for point clouds and images. Our model demonstrates the capability to generate meshes with intricate details and accurate topology, achieving SoTA performance on mesh generation and reaching the level for direct product usage., Comment: Homepage: https://whaohan.github.io/bpt , Code: https://github.com/whaohan/bpt

While 3D generative models have greatly improved artists' workflows, the existing diffusion models for 3D generation suffer from slow generation and poor generalization. To address this issue, we propose a two-stage approach named Hunyuan3D-1.0 including a lite version and a standard version, that both support text- and image-conditioned generation. In the first stage, we employ a multi-view diffusion model that efficiently generates multi-view RGB in approximately 4 seconds. These multi-view images capture rich details of the 3D asset from different viewpoints, relaxing the tasks from single-view to multi-view reconstruction. In the second stage, we introduce a feed-forward reconstruction model that rapidly and faithfully reconstructs the 3D asset given the generated multi-view images in approximately 7 seconds. The reconstruction network learns to handle noises and in-consistency introduced by the multi-view diffusion and leverages the available information from the condition image to efficiently recover the 3D structure. Our framework involves the text-to-image model, i.e., Hunyuan-DiT, making it a unified framework to support both text- and image-conditioned 3D generation. Our standard version has 3x more parameters than our lite and other existing model. Our Hunyuan3D-1.0 achieves an impressive balance between speed and quality, significantly reducing generation time while maintaining the quality and diversity of the produced assets., Comment: Technical Report; 3D Generation

In this paper, we introduce Hunyuan-Large, which is currently the largest open-source Transformer-based mixture of experts model, with a total of 389 billion parameters and 52 billion activation parameters, capable of handling up to 256K tokens. We conduct a thorough evaluation of Hunyuan-Large's superior performance across various benchmarks including language understanding and generation, logical reasoning, mathematical problem-solving, coding, long-context, and aggregated tasks, where it outperforms LLama3.1-70B and exhibits comparable performance when compared to the significantly larger LLama3.1-405B model. Key practice of Hunyuan-Large include large-scale synthetic data that is orders larger than in previous literature, a mixed expert routing strategy, a key-value cache compression technique, and an expert-specific learning rate strategy. Additionally, we also investigate the scaling laws and learning rate schedule of mixture of experts models, providing valuable insights and guidances for future model development and optimization. The code and checkpoints of Hunyuan-Large are released to facilitate future innovations and applications. Codes: https://github.com/Tencent/Hunyuan-Large Models: https://huggingface.co/tencent/Tencent-Hunyuan-Large, Comment: 17 pages, 4 Figures

As Large Language Models (LLMs) grow increasingly adept at managing complex tasks, the evaluation set must keep pace with these advancements to ensure it remains sufficiently discriminative. Item Discrimination (ID) theory, which is widely used in educational assessment, measures the ability of individual test items to differentiate between high and low performers. Inspired by this theory, we propose an ID-induced prompt synthesis framework for evaluating LLMs to ensure the evaluation set can continually update and refine according to model abilities. Our data synthesis framework prioritizes both breadth and specificity. It can generate prompts that comprehensively evaluate the capabilities of LLMs while revealing meaningful performance differences between models, allowing for effective discrimination of their relative strengths and weaknesses across various tasks and domains. To produce high-quality data, we incorporate a self-correct mechanism into our generalization framework, and develop two models to predict prompt discrimination and difficulty score to facilitate our data synthesis framework, contributing valuable tools to evaluation data synthesis research. We apply our generated data to evaluate five SOTA models. Our data achieves an average score of 51.92, accompanied by a variance of 10.06. By contrast, previous works (i.e., SELF-INSTRUCT and WizardLM) obtain an average score exceeding 67, with a variance below 3.2. The results demonstrate that the data generated by our framework is more challenging and discriminative compared to previous works. We will release a dataset of over 3,000 carefully crafted prompts to facilitate evaluation research of LLMs., Comment: NeurIPS 2024

Identifying IoT devices is crucial for network monitoring, security enforcement, and inventory tracking. However, most existing identification methods rely on deep packet inspection, which raises privacy concerns and adds computational complexity. More importantly, existing works overlook the impact of wireless channel dynamics on the accuracy of layer-2 features, thereby limiting their effectiveness in real-world scenarios. In this work, we define and use the latency of specific probe-response packet exchanges, referred to as "device latency," as the main feature for device identification. Additionally, we reveal the critical impact of wireless channel dynamics on the accuracy of device identification based on device latency. Specifically, this work introduces "accumulation score" as a novel approach to capturing fine-grained channel dynamics and their impact on device latency when training machine learning models. We implement the proposed methods and measure the accuracy and overhead of device identification in real-world scenarios. The results confirm that by incorporating the accumulation score for balanced data collection and training machine learning algorithms, we achieve an F1 score of over 97% for device identification, even amidst wireless channel dynamics, a significant improvement over the 75% F1 score achieved by disregarding the impact of channel dynamics on data collection and device latency.

We present Hunyuan-DiT, a text-to-image diffusion transformer with fine-grained understanding of both English and Chinese. To construct Hunyuan-DiT, we carefully design the transformer structure, text encoder, and positional encoding. We also build from scratch a whole data pipeline to update and evaluate data for iterative model optimization. For fine-grained language understanding, we train a Multimodal Large Language Model to refine the captions of the images. Finally, Hunyuan-DiT can perform multi-turn multimodal dialogue with users, generating and refining images according to the context. Through our holistic human evaluation protocol with more than 50 professional human evaluators, Hunyuan-DiT sets a new state-of-the-art in Chinese-to-image generation compared with other open-source models. Code and pretrained models are publicly available at github.com/Tencent/HunyuanDiT, Comment: Project Page: https://dit.hunyuan.tencent.com/

The emerging Self-Sovereign Identity (SSI) techniques, such as Decentralized Identifiers (DIDs) and Verifiable Credentials (VCs), move control of digital identity from conventional identity providers to individuals and lay down the foundation for people, organizations, and things establishing rich digital relationship. The existing applications of SSI mainly focus on creating person-to-person and person-to-service relationships, whereas person-to-device and device-to-device interactions have been largely overlooked. In this paper, we close this gap by identifying a number of key challenges of applying SSI to the Internet of Things (IoT) and providing a comprehensive taxonomy and usage of VCs in the IoT context with respect to their validity period, trust and interoperability level, and scope of usage. The life-cycle management of VCs as well as various optimization techniques for realizing SSI in IoT environments are also addressed in great detail. This work is a noteworthy step towards massive adoption of SSI for securing existing and future IoT applications in practice.

The rapid advancements in generative AI and large language models (LLMs) have opened up new avenues for producing synthetic data, particularly in the realm of structured tabular formats, such as product reviews. Despite the potential benefits, concerns regarding privacy leakage have surfaced, especially when personal information is utilized in the training datasets. In addition, there is an absence of a comprehensive evaluation framework capable of quantitatively measuring the quality of the generated synthetic data and their utility for downstream tasks. In response to this gap, we introduce SynEval, an open-source evaluation framework designed to assess the fidelity, utility, and privacy preservation of synthetically generated tabular data via a suite of diverse evaluation metrics. We validate the efficacy of our proposed framework - SynEval - by applying it to synthetic product review data generated by three state-of-the-art LLMs: ChatGPT, Claude, and Llama. Our experimental findings illuminate the trade-offs between various evaluation metrics in the context of synthetic data generation. Furthermore, SynEval stands as a critical instrument for researchers and practitioners engaged with synthetic tabular data,, empowering them to judiciously determine the suitability of the generated data for their specific applications, with an emphasis on upholding user privacy., Comment: 10 pages, 1 figure, 4 tables

Privacy-Preserving Neural Networks (PPNN) are advanced to perform inference without breaching user privacy, which can serve as an essential tool for medical diagnosis to simultaneously achieve big data utility and privacy protection. As one of the key techniques to enable PPNN, Fully Homomorphic Encryption (FHE) is facing a great challenge that homomorphic operations cannot be easily adapted for non-linear activation calculations. In this paper, batch-oriented element-wise data packing and approximate activation are proposed, which train linear low-degree polynomials to approximate the non-linear activation function - ReLU. Compared with other approximate activation methods, the proposed fine-grained, trainable approximation scheme can effectively reduce the accuracy loss caused by approximation errors. Meanwhile, due to element-wise data packing, a large batch of images can be packed and inferred concurrently, leading to a much higher utility ratio of ciphertext slots. Therefore, although the total inference time increases sharply, the amortized time for each image actually decreases, especially when the batch size increases. Furthermore, knowledge distillation is adopted in the training process to further enhance the inference accuracy. Experiment results show that when ciphertext inference is performed on 4096 input images, compared with the current most efficient channel-wise method, the inference accuracy is improved by 1.65%, and the amortized inference time is reduced by 99.5%.

The Diffusion model has a strong ability to generate wild images. However, the model can just generate inaccurate images with the guidance of text, which makes it very challenging to directly apply the text-guided generative model for virtual try-on scenarios. Taking images as guiding conditions of the diffusion model, this paper proposes a brand new personalized virtual try-on model (PE-VITON), which uses the two stages (shape control and texture guidance) to decouple the clothing attributes. Specifically, the proposed model adaptively matches the clothing to human body parts through the Shape Control Module (SCM) to mitigate the misalignment of the clothing and the human body parts. The semantic information of the input clothing is parsed by the Texture Guided Module (TGM), and the corresponding texture is generated by directional guidance. Therefore, this model can effectively solve the problems of weak reduction of clothing folds, poor generation effect under complex human posture, blurred edges of clothing, and unclear texture styles in traditional try-on methods. Meanwhile, the model can automatically enhance the generated clothing folds and textures according to the human posture, and improve the authenticity of virtual try-on. In this paper, qualitative and quantitative experiments are carried out on high-resolution paired and unpaired datasets, the results show that the proposed model outperforms the state-of-the-art model.

In this paper, we develop a deep learning-based bandwidth allocation policy that is: 1) scalable with the number of users and 2) transferable to different communication scenarios, such as non-stationary wireless channels, different quality-of-service (QoS) requirements, and dynamically available resources. To support scalability, the bandwidth allocation policy is represented by a graph neural network (GNN), with which the number of training parameters does not change with the number of users. To enable the generalization of the GNN, we develop a hybrid-task meta-learning (HML) algorithm that trains the initial parameters of the GNN with different communication scenarios during meta-training. Next, during meta-testing, a few samples are used to fine-tune the GNN with unseen communication scenarios. Simulation results demonstrate that our HML approach can improve the initial performance by $8.79\%$, and sampling efficiency by $73\%$, compared with existing benchmarks. After fine-tuning, our near-optimal GNN-based policy can achieve close to the same reward with much lower inference complexity compared to the optimal policy obtained using iterative optimization.

This paper designs a graph neural network (GNN) to improve bandwidth allocations for multiple legitimate wireless users transmitting to a base station in the presence of an eavesdropper. To improve the privacy and prevent eavesdropping attacks, we propose a user scheduling algorithm to schedule users satisfying an instantaneous minimum secrecy rate constraint. Based on this, we optimize the bandwidth allocations with three algorithms namely iterative search (IvS), GNN-based supervised learning (GNN-SL), and GNN-based unsupervised learning (GNN-USL). We present a computational complexity analysis which shows that GNN-SL and GNN-USL can be more efficient compared to IvS which is limited by the bandwidth block size. Numerical simulation results highlight that our proposed GNN-based resource allocations can achieve a comparable sum secrecy rate compared to IvS with significantly lower computational complexity. Furthermore, we observe that the GNN approach is more robust to uncertainties in the eavesdropper's channel state information, especially compared with the best channel allocation scheme.

Large language models (LLMs) have shown impressive capabilities across various natural language tasks. However, evaluating their alignment with human preferences remains a challenge. To this end, we propose a comprehensive human evaluation framework to assess LLMs' proficiency in following instructions on diverse real-world tasks. We construct a hierarchical task tree encompassing 7 major areas covering over 200 categories and over 800 tasks, which covers diverse capabilities such as question answering, reasoning, multiturn dialogue, and text generation, to evaluate LLMs in a comprehensive and in-depth manner. We also design detailed evaluation standards and processes to facilitate consistent, unbiased judgments from human evaluators. A test set of over 3,000 instances is released, spanning different difficulty levels and knowledge domains. Our work provides a standardized methodology to evaluate human alignment in LLMs for both English and Chinese. We also analyze the feasibility of automating parts of evaluation with a strong LLM (GPT-4). Our framework supports a thorough assessment of LLMs as they are integrated into real-world applications. We have made publicly available the task tree, TencentLLMEval dataset, and evaluation methodology which have been demonstrated as effective in assessing the performance of Tencent Hunyuan LLMs. By doing so, we aim to facilitate the benchmarking of advances in the development of safe and human-aligned LLMs.

The rapid development of intelligent transportation systems and connected vehicles has highlighted the need for secure and efficient key management systems (KMS). In this paper, we introduce VDKMS (Vehicular Decentralized Key Management System), a novel Decentralized Key Management System designed specifically as an infrastructure for Cellular Vehicular-to-Everything (V2X) networks, utilizing a blockchain-based approach. The proposed VDKMS addresses the challenges of secure communication, privacy preservation, and efficient key management in V2X scenarios. It integrates blockchain technology, Self-Sovereign Identity (SSI) principles, and Decentralized Identifiers (DIDs) to enable secure and trustworthy V2X applications among vehicles, infrastructures, and networks. We first provide a comprehensive overview of the system architecture, components, protocols, and workflows, covering aspects such as provisioning, registration, verification, and authorization. We then present a detailed performance evaluation, discussing the security properties and compatibility of the proposed solution, as well as a security analysis. Finally, we present potential applications in the vehicular ecosystem that can leverage the advantages of our approach., Comment: 6 pages, 6 figures, accepted by IEEE Globecom 2023

The development of machine learning techniques enables us to construct surrogate models from data of direct numerical simulations, which has important implications for modeling complex physical systems. In this paper, based on the output from 1D Vlasov-Ampere simulations, we adopt the Fourier Neural Operator (FNO) to build surrogate models of Landau fluid closure for multi-moment fluid equations from kinetic simulation data. The trained FNO is able to obtain the heat flux using electron density as input, in agreement with the true value from kinetic simulations. We compare the physical quantities obtained using the FNO and Multilayer Perceptron (MLP) architectures, and found that the results of FNO are significantly better than that of MLP., Comment: 4 pages, 5 figures, to appear in "International Applied Computational Electromagnetics Society Symposium (ACES)" by IEEE

With the massive amount of digital data generated everyday, transactions of digital goods become a trend. One of the essential requirements for such transactions is fairness, which is defined as that both of the seller and the buyer get what they want, or neither. Current fair trade protocols generally involve a trusted third-party (TTP), which achieves fairness by heavily relying on the TTP's behaviors and the two parties' trust in the TTP. With the emergence of Blockchain, its decentralization and transparency make it a very good candidate to replace the TTP. In this work, we attempt to design a secure and fair protocol for digital goods transactions through smart contracts on Blockchain. To ensure security of the digital goods, we propose an advanced passive proxy re-encryption (PRE) scheme, which enables smart contracts to transfer the decryption right to a buyer after receiving his/her payment. Furthermore, based on smart contracts and the proposed passive PRE scheme, a fair trade protocol for digital goods transactions is proposed, whose fairness is guaranteed by the arbitration protocol. The proposed protocol supports Ciphertext publicity and repeatable sale, while involving less number of interactions. Comprehensive experiment results validate the feasibility and effectiveness of the proposed protocol.

Multi-signature aggregates signatures from multiple users on the same message into a joint signature, which is widely applied in blockchain to reduce the percentage of signatures in blocks and improve the throughput of transactions. The $k$-sum attacks are one of the major challenges to design secure multi-signature schemes. In this work, we address $k$-sum attacks from a novel angle by defining a Public Third Party (PTP), which is an automatic process that can be verifiable by the public and restricts the signing phase from continuing until receiving commitments from all signers. Further, a two-round multi-signature scheme MEMS with PTP is proposed, which is secure based on discrete logarithm assumption in the random oracle model. As each signer communicates directly with the PTP instead of other co-signers, the total amount of communications is significantly reduced. In addition, as PTP participates in the computation of the aggregation and signing algorithms, the computation cost left for each signer and verifier remains the same as the basis Schnorr signature. To the best of our knowledge, this is the maximum efficiency that a Schnorr-based multi-signature scheme can achieve. Further, MEMS is applied in blockchain platform, e.g., Fabric, to improve the transaction efficiency.

Digital signature is a major component of transactions on Blockchain platforms, especially in enterprise Blockchain platforms, where multiple signatures from a set of peers need to be produced to endorse a transaction. However, such process is often complex and time-consuming. Multi-signature, which can improve transaction efficiency by having a set of signers cooperate to produce a joint signature, has attracted extensive attentions. In this work, we propose two multi-signature schemes, GMS and AGMS, which are proved to be more secure and efficient than state-of-the-art multi-signature schemes. Besides, we implement the proposed schemes in a real Enterprise Blockchain platform, Fabric. Experiment results show that the proposed AGMS scheme helps achieve the goal of high transaction efficiency, low storage complexity, as well as high robustness against rogue-key attacks and k-sum problem attacks.

In this paper, we aim to improve the Quality-of-Service (QoS) of Ultra-Reliability and Low-Latency Communications (URLLC) in interference-limited wireless networks. To obtain time diversity within the channel coherence time, we first put forward a random repetition scheme that randomizes the interference power. Then, we optimize the number of reserved slots and the number of repetitions for each packet to minimize the QoS violation probability, defined as the percentage of users that cannot achieve URLLC. We build a cascaded Random Edge Graph Neural Network (REGNN) to represent the repetition scheme and develop a model-free unsupervised learning method to train it. We analyze the QoS violation probability using stochastic geometry in a symmetric scenario and apply a model-based Exhaustive Search (ES) method to find the optimal solution. Simulation results show that in the symmetric scenario, the QoS violation probabilities achieved by the model-free learning method and the model-based ES method are nearly the same. In more general scenarios, the cascaded REGNN generalizes very well in wireless networks with different scales, network topologies, cell densities, and frequency reuse factors. It outperforms the model-based ES method in the presence of the model mismatch., Comment: Submitted to IEEE journal for possible publication

Resource sharing among users serves as the foundation of cloud computing, which, however, may also cause vulnerabilities to diverse co-residence attacks launched by malicious virtual machines (VM) residing in the same physical server with the victim VMs. In this paper, we aim to defend against such co-residence attacks through a secure, workload-balanced, and energy-efficient VM allocation strategy. Specifically, we model the problem as an optimization problem by quantifying and minimizing three key factors: (1) the security risks, (2) the power consumption and (3) the unbalanced workloads among different physical servers. Furthermore, this work considers a realistic environmental setting by assuming a random number of VMs from different users arriving at random timings, which requires the optimization solution to be continuously evolving. As the optimization problem is NP-hard, we propose to first cluster VMs in time windows, and further adopt the Ant Colony Optimization (ACO) algorithm to identify the optimal allocation strategy for each time window. Comprehensive experimental results based on real world cloud traces validates the effectiveness of the proposed scheme.

IPv6 Routing Protocol for Low Power and Lossy Networks (RPL) is an essential routing protocol to enable communications for IoT networks with low power devices. RPL uses an objective function and routing constraints to find an optimized routing path for each node in the network. However, recent research has shown that topological attacks, such as selective forwarding attacks, pose great challenges to the secure routing of IoT networks. Many conventional secure routing solutions, on the other hand, are computationally heavy to be directly applied in resource-constrained IoT networks. There is an urgent need to develop lightweight secure routing solutions for IoT networks. In this paper, we first design and implement a series of advanced selective forwarding attacks from the attack perspective, which can flexibly select the type and percentage of forwarding packets in an energy efficient way, and even bad-mouth other innocent nodes in the network. Experiment results show that the proposed attacks can maximize the attack consequences (i.e. number of dropped packets) while maintaining undetected. Moreover, we propose a lightweight trust-based defense solution to detect and eliminate malicious selective forwarding nodes from the network. The results show that the proposed defense solution can achieve high detection accuracy with very limited extra energy usage (i.e. 3.4%).