Jie Wang

An Iterative Polishing Framework based on Quality Aware Masked Language Model for Chinese Poetry Generation

2019 · arXiv (Cornell University)

Owing to its unique literal and aesthetical characteristics, automatic generation of Chinese poetry is still challenging in Artificial Intelligence, which can hardly be straightforwardly realized by end-to-end methods. In this paper, we propose a novel …

On Explainability of Graph Neural Networks via Subgraph Explorations

2021 · arXiv (Cornell University)

We consider the problem of explaining the predictions of graph neural networks (GNNs), which otherwise are considered as black boxes. Existing methods invariably focus on explaining the importance of graph nodes or edges but ignore …

A Blockchain-Based Public Auditing Protocol with Self-Certified Public Keys for Cloud Data

2021 · Security and Communication Networks

Cloud storage can provide a way to effectively store and manage big data. However, due to the separation of data ownership and management, it is difficult for users to check the integrity of data in …

Adversarially Learning Disentangled Speech Representations for Robust Multi-Factor Voice Conversion

2021

Factorizing speech as disentangled speech representations is vital to achieve highly controllable style transfer in voice conversion (VC).Conventional speech representation learning methods in VC only factorize speech as speaker and content, lacking controllability on other …

Democratizing Domain-Specific Computing

2022 · arXiv (Cornell University)

In the past few years, domain-specific accelerators (DSAs), such as Google's Tensor Processing Units, have shown to offer significant performance and energy efficiency over general-purpose CPUs. An important question is whether typical software developers can …

Event-triggered sequential fusion filters for multi-sensor systems with multiplicative noises

2022 · 2022 37th Youth Academic Annual Conference of Chinese Association of Automation (YAC)

This paper combines an event-triggered mechanism with a multiplicative noise system to reduce the communication congestion and the calculation workload of the networked fusion center, and proposes an event-triggered Sequence Covariance Intersection and Sequence Inverse …

Local Texture Complexity Guided Adversarial Attack

2023

Extensive research revealed that deep neural networks are vulnerable to adversarial examples. In addition, recent studies have demonstrated that convolutional neural networks tend to recognize the texture (high-frequency components) rather than the shape (low-frequency components) …

Content Significance Distribution of Sub-Text Blocks in Articles and Its Application to Article-Organization Assessment

2023 · arXiv (Cornell University)

We explore how to capture the significance of a sub-text block in an article and how it may be used for text mining tasks. A sub-text block is a sub-sequence of sentences in the article. …

Multi-Augmentation-Based Contrastive Learning for Semi-Supervised Learning

2024 · Algorithms

Semi-supervised learning has been proven to be effective in utilizing unlabeled samples to mitigate the problem of limited labeled data. Traditional semi-supervised learning methods generate pseudo-labels for unlabeled samples and train the classifier using both …

Capturing the Spectrum of Social Media Conflict: A Novel Multi-objective Classification Model

2024

Social media has emerged as a widespread phenomenon, with numerous users engaging in observing, creating, and distributing content. The growing content has led to user conflicts, encompassing bullying, aggression, harassment, and threats. Consequently, recent research …

Identification of Usage Types for New Energy Vehicles Based on Big Data Analytics

2025

Accurately determining the true usage types of new energy vehicles (NEVs) is crucial for insurance pricing and second-hand market valuation as their use grows. Effective decision-making is hampered by discrepancies between registered and actual use. …

StarVC: A Unified Auto-Regressive Framework for Joint Text and Speech Generation in Voice Conversion

2025 · arXiv (Cornell University)

Voice Conversion (VC) modifies speech to match a target speaker while preserving linguistic content. Traditional methods usually extract speaker information directly from speech while neglecting the explicit utilization of linguistic content. Since VC fundamentally involves …

Accurate and Scalable Graph Neural Networks via Message Invariance

2025 · arXiv (Cornell University)

Message passing-based graph neural networks (GNNs) have achieved great success in many real-world applications. For a sampled mini-batch of target nodes, the message passing process is divided into two parts: message passing between nodes within …

S2-Boost: Synergistic Semantic Boosting for Coarse-to-Fine Ensemble Learning

2026 · Proceedings of the AAAI Conference on Artificial Intelligence

Neuroscientific evidence reveals that human visual recognition is not an instantaneous event but a hierarchical process, where the brain constructs a holistic perception by progressively integrating simple features like edges or texture into complex scenes. …