Lightweight Complementary-Cue Fusion for Robust Video Face Forgery Detection
This paper proposes two lightweight face forgery detectors, LFWS and LFWL, built on Xception (21.9M params) by adding a fusion module with only 292 extra parameters. They combine wavelet-denoised features with phase spectrum or local binary patterns, boosting AUC by 3.8% and 4.4% on FaceForensics++ and DFDC-Preview, respectively, outperforming larger models like F3Net and SRM across eight benchmarks.
Article intelligence
Key points
- LFWS and LFWL add only 292 parameters to Xception, keeping total at 21.9M, smaller than F3Net (22.5M) and less than half of SRM (55.3M).
- AUC improves from 74.8% to 78.6% on FaceForensics++ and from 70.5% to 74.9% on DFDC-Preview, gains of 3.8% and 4.4%.
- Consistently outperform F3Net, SRM, and SPSL on eight public benchmarks without extra data or test-time augmentation.
- Results show that carefully paired handcrafted features via lightweight fusion offer competitive robustness at lower cost, challenging scale-driven designs.
Why it matters
This matters because LFWS and LFWL add only 292 parameters to Xception, keeping total at 21.9M, smaller than F3Net (22.5M) and less than half of SRM (55.3M).
Technical impact
May affect model selection, inference cost, product capability, and evaluation benchmarks.
[2605.29092] Lightweight Complementary-Cue Fusion for Robust Video Face Forgery Detection
[Submitted on 27 May 2026]
Title:Lightweight Complementary-Cue Fusion for Robust Video Face Forgery Detection
View a PDF of the paper titled Lightweight Complementary-Cue Fusion for Robust Video Face Forgery Detection, by Sunghwan Baek and 3 other authors
View PDF HTML (experimental)
Abstract:Current face video forgery detectors use wide or dual-stream backbones. We show that a single, lightweight fusion of two handcrafted cues can achieve higher accuracy with a much smaller model. Based on the Xception baseline model (21.9 million parameters), we build two detectors: LFWS, which adds a 1x1 convolution to combine a low-frequency Wavelet-Denoised Feature (WDF) with a phase-spectrum channel derived from Spatial-Phase Shallow Learning (SPSL), and LFWL, which merges WDF with Local Binary Patterns (LBP) in the same way. This extra module adds only 292 parameters, keeping the total at 21.9 million, smaller than F3Net (22.5 million) and less than half the size of SRM (55.3 million). Even with this minimal overhead, the fused models increase the average area under the curve (AUC) from 74.8% to 78.6% on FaceForensics++ and from 70.5% to 74.9% on DFDC-Preview, gains of 3.8% and 4.4% over the Xception baseline. They also consistently outperform F3Net, SRM, and SPSL in eight public benchmarks, without extra data or test-time augmentation. These results show that carefully paired, handcrafted features, combined through the lightweight fusion block, can provide competitive robustness at a significantly lower cost than comparable frequency-based detectors. Our findings suggest a need to reevaluate scale-driven design choices in face video forgery detection.
Comments: 13 pages, 6 figures, 3 tables
Subjects:
Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Multimedia (cs.MM)
ACM classes: I.4.9; I.5.4
Cite as: arXiv:2605.29092 [cs.CV]
(or arXiv:2605.29092v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2605.29092
arXiv-issued DOI via DataCite (pending registration)
Submission history
From: Sunghwan Baek [view email] [v1] Wed, 27 May 2026 20:51:04 UTC (2,382 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled Lightweight Complementary-Cue Fusion for Robust Video Face Forgery Detection, by Sunghwan Baek and 3 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.CV
new | recent | 2026-05
Change to browse by:
cs cs.LG cs.MM
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
Loading...
Data provided by:
Bibliographic Tools
Bibliographic and Citation Tools
Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media
Code, Data and Media Associated with this Article
alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos
Demos
Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers
Recommenders and Search Tools
Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
Author
Venue
Institution
Topic
About arXivLabs
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)