Image Generation Evaluation: A Comprehensive Survey of Human and Automatic Evaluations

General

Towards the first principles of explaining DNNs: interactions explain the learning dynamics

Published On : 2025-10-17

In the field of explainable artificial intelligence (XAI), deep neural networks (DNNs) face a critical challenge: their superior performance in tasks like image classification and natural language processing is accompanied by opaque decision-making logic, especially in high-risk scenarios such as medical diagnosis and autonomous driving.

#General

General

Risks and Mechanistic Insights Into Arsenic-Enhanced Iodination of Bisphenol F in Brassica Chinensis L.

Published On : 2025-10-17

This study highlights the combined risks of arsenic and bisphenol contamination, revealing arsenic’s role in enhancing bisphenol iodination and toxicity in plants.

#General

General

Output Difference Feedback and System Benefit Control Based Dynamic Heterogeneous Redundancy Architecture

Published On : 2025-10-17

With the continuous development of cyberspace, traditional passive defense technologies are difficult to effectively respond to complex cybersecurity threats due to their reactive and specific characteristics. Mimic active defense technology, as a typical active defense method, uses a dynamic heterogeneous redundancy (DHR) architecture to disrupt attack routes and reduce the probability of successful attacks. However, current DHR architecture approaches often ignore the adaptability of the adjudication mechanism in complex and variable network environments, focusing mainly on system security…

#General

General

Image Generation Evaluation: A Comprehensive Survey of Human and Automatic Evaluations

Published On : 2025-10-17

In the field of image generation, where models have achieved remarkable progress and are widely applied in medicine, fashion, and media, evaluation methods have failed to keep pace with model development–creating a critical disconnect.

#General