Detecting AI-Generated Videos with Spiking Neural Networks

Minsuk Jang, Yujin Yang, Heeseon Kim, Younghun Kim, Chang-Ick Kim · 2026

Read Detecting AI-Generated Videos with Spiking Neural Networks by Minsuk Jang, Yujin Yang, Heeseon Kim, Younghun Kim, Chang-Ick Kim (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-05895.txt.

Description

AI-generated video now looks real frame by frame, so detectors have to read the motion between frames. Minsuk Jang and four co-authors use spiking neural networks for the job, on the GenVidBench benchmark. They find two signatures earlier detectors do not jointly use, beginning with smoother frame-to-frame temporal residuals in generated footage at the pixel level.

Detectors built on single cues or fixed descriptors fail sharply when the generator changes. Spiking networks, by nature sensitive to timing, are matched to the surviving evidence. The work aims at detection that holds up against unseen generators.

Catalog record

CREATORMinsuk Jang, Yujin Yang, Heeseon Kim, Younghun Kim, Chang-Ick Kim
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE88 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.05895

Browse the full library · About the Tower of Babel