Tag: Information Security

Yuanli Technology Paper Accepted into the IEEE Database | Deepfake Liveness Detection Technology Gains International Recognition

With the large-scale adoption of artificial intelligence, AIGC (AI-Generated Content) is profoundly transforming content creation. However, while convenience and efficiency continue to improve, the security risks introduced by AIGC are also escalating rapidly. Among them, deepfake technology has become a key driver of online fraud, misinformation, and identity impersonation. Increasingly realistic forged facial images and videos are placing unprecedented pressure on enterprise identity verification, risk control processes, and information security systems.

Against this backdrop, Yuanli Technology has continued to focus on deepfake detection research. Recently, a research paper on AIGC deepfake liveness detection, authored by the FinAuth algorithm team, was successfully accepted by the 2025 International Joint Conference on Neural Networks (IJCNN) and officially indexed by IEEE.

IJCNN is one of the most authoritative international conferences in the field of neural networks and computational intelligence. This acceptance not only represents international recognition of our research achievements, but also demonstrates our company’s continuous innovation and technical strength in deepfake detection technologies.

Escalating Deepfake Risks: Traditional Detection Technologies Struggle with Emerging Attacks

In recent years, the rapid development of GANs, diffusion models, and other generative technologies has significantly improved the quality of forged images. Along with this progress, two major challenges have emerged across the industry:

1. Strong Dependence on Specific Forgery Techniques

Many existing detection methods perform well only on specific types of forged data. When faced with unknown or newly emerging generation models, detection accuracy drops sharply. This means enterprises often remain passive in responding to new deepfake attacks, requiring continuous model retraining and adaptation to different forgery methods, making it difficult to achieve long-term and stable security protection.

2. Difficulty in Identifying Advanced Forgeries

As forged images become increasingly realistic, detection methods that rely on explicit features such as facial expressions or skin tone are gradually losing effectiveness. Many advanced deepfakes achieve near-perfect visual disguise, making it difficult for traditional algorithms to capture the subtle anomalies hidden within fine-grained texture details.

These challenges significantly increase the risks associated with identity fraud and verification processes, driving an urgent industry need for more generalizable and robust detection methods.

A Frequency-Domain Self-Supervised Deepfake Detection Framework Proposed and Accepted by IJCNN 2025

The IEEE-accepted paper introduces a forged face detection technology based on frequency-domain self-supervised learning. The core idea of this approach is to analyze an image’s frequency information by decomposing it into high-frequency and low-frequency components, with particular emphasis on high-frequency components, which typically represent fine textures and edge details within images.

Through this method, the algorithm can more accurately identify abnormal features in forged images, thereby improving overall detection accuracy.

Overview of proposed FreSSL method

Compared with existing methods, the proposed approach demonstrates significant advantages in the following aspects:

1. Focusing on high-frequency image information to reduce interference from facial expressions and skin tone, thereby improving accuracy

Traditional forged image detection methods may be affected by factors such as facial expressions or skin tone in images. In contrast, this approach leverages self-supervised learning to focus on the reconstruction of high-frequency information, concentrating on subtle details that are difficult to perceive. By avoiding reliance on prominent visual features, the method is not misled by obvious appearance cues, thus improving detection accuracy.

2. Pretrained on real facial data, exhibiting strong generalization across different diffusion-based forgeries

By transforming images from the spatial domain to the frequency domain, this method enables more effective differentiation between real facial images and forged ones. Its key advantage lies in pretraining on real facial datasets, allowing the model to maintain strong generalization performance even when encountering previously unseen forged data generated by different diffusion models.

AIGC technologies bring innovation and efficiency to society, but they also introduce new security challenges. Our company remains committed to independent technological research combined with academic openness, continuously investing in areas such as deepfake detection and defense algorithms.

The acceptance of this paper by IEEE represents international recognition of our technical capabilities and marks an important step in promoting industry-wide efforts to address AIGC-related risks. Looking ahead, we will continue to explore more advanced deepfake detection solutions, safeguard digital identity security, and promote the healthy and responsible development of AI technologies.

This article provides only a summary of the core research content. Full technical details, experimental design, and result analysis are available on IEEE Xplore:

https://ieeexplore.ieee.org/document/11227294

January 22, 2026
FinAuth: Financial-Grade Identity Verification — Fortifying Intelligent Security, Empowering Business Growth
In the wave of digital transformation, identity fraud and AI-driven synthetic attacks are evolving at an alarming pace — causing financial loss and security risks across multiple industries.

Facing the dual challenge of security protection and business growth, FinAuth leverages a dual-engine “Edge + Cloud” liveness detection framework and multimodal AI vision models to build a financial-grade intelligent security defense, becoming a pivotal key to solving the current growth dilemma.

FinAuth Financial-Grade Identity Verification Services

The FinAuth cloud-based identity authentication platform is a financial-grade KYC verification platform based on facial recognition.

It integrates three core capabilities: ID OCR, liveness detection, and face comparison. With the power of large AI models, FinAuth delivers strong anti-attack capabilities, high success rates, and adaptability to complex lighting conditions — helping enterprises improve operational efficiency, reduce fraud risks, and increase business performance.

At present, the service has been successfully applied in financial, mobility, and social entertainment sectors, intercepting over one billion fraudulent attacks, and becoming a reliable partner for enterprises in risk prevention and operational optimization.

FinAuth Financial-Grade Identity Verification Services

Why Choose FinAuth

1. Multiple Liveness Detection Modes — Flexibly Adapt to Different Business Scenarios

FinAuth provides liveness detection modes such as flash, active, 3D liveness and intention, forming a dynamically configurable defense matrix.

Through randomized color sequence validation, spatial movement detection, and multi-frame action analysis, FinAuth precisely defends against more than ten types of attacks such as AI-generated faces, screen replay, and video injection.

Enterprises can flexibly combine detection methods based on risk level — for example, low-risk scenarios can adopt simplified flash liveness with only 3 seconds to complete verification; high-risk scenarios may add 3D or intention-based liveness to enhance security through dual confirmation.

This flexible liveness detection solution meets financial-grade protection requirements while ensuring user experience.

2. Proactive Defense System — Building an Intelligent Security Firewall
- Introducing vision large models for rapid and accurate risk judgment: Based on extensive industry practices and historical attack behaviors, the vision large model can quickly determine risky requests and improve efficiency and accuracy of risk identification.
- Multi-dimensional risk control strategies — fully enhancing system defense capability: By combining AI vision models, dynamic blacklists, and anti-injection strategies, FinAuth builds a multi-layer risk control barrier, significantly improving attack detection rates and ensuring uncompromised security.
- Innovative random factor detection to defend against device-based attacks: By integrating liveness detection with random factor verification, FinAuth can accurately determine whether the video is truly captured by the SDK, effectively preventing video injection attacks and making device-based attacks impossible to hide.
3. High Success Rate Engine — Experience Optimization and Business Growth
- Excellent interaction experience, improving user satisfaction: With a futuristic hexagonal design and privacy-optimized interface, users enjoy a cleaner and smoother experience.
- Efficient comparison and fast response: Fast API response and accurate algorithm recognition enable quick completion of identity verification, significantly enhancing user experience and reducing waiting time.
- Stable performance under complex lighting conditions: The robust liveness algorithm operates stably under strong light, backlight, and side light, maintaining an average pass rate of over 99%, ensuring efficient operation across all business scenarios.
Focusing on Industry Pain Points — Securing Core Business

Finance Industry: Strengthening Risk Control Foundations and Safeguarding Funds

In banking, insurance, and securities scenarios, FinAuth provides full-process protection for high-risk operations. From real-name verification during online account opening, to live person confirmation in loan approval, to secondary identity verification in mobile payments and large transfers — FinAuth uses dynamic liveness detection and AI anti-fraud models to accurately block face-swapping and synthetic video attacks, enabling enterprises to achieve breakthroughs in both security and efficiency.

Mobility Industry: Real-Name Management and Improved Service Trust

For shared mobility platforms, FinAuth helps enterprises build a trustworthy transportation ecosystem.

During driver identity verification, distance detection and screen replay recognition eliminate false identity registration and prevent fraudulent activities.

After adoption by a leading mobility platform, complaints related to driver identity impersonation dropped by 85%, while customer satisfaction and platform performance significantly improved.

Social Entertainment Industry: Clean Ecosystem and Better User Experience

In live streaming, gaming, and social platforms, FinAuth protects content compliance and user safety.

Live real-person authentication filters out malicious accounts before streaming; during high-value tipping or virtual asset transactions, intention-based liveness detection confirms true user consent and reduces dispute rates; in anti-addiction scenarios, AI algorithms accurately detect face-swapping and ensure regulatory compliance.

Take Action Now — Enable Both Security and Growth

To help enterprises respond quickly to risk challenges, FinAuth now offers free technical consultation and five liveness detection demo experiences.

Financial customers can also apply for customized solutions and obtain exclusive risk control strategy optimization services.

FinAuth — Reshaping Trust with Technology, Turning Every Identity Verification into the Starting Point of Business Growth.
January 22, 2026
The Challenge of Identity Verification in Global Finance: Finding the “Golden Balance” Between Security and User Experience

As financial institutions accelerate their global expansion, identity verification has evolved from a simple compliance procedure into a core security foundation that runs through the entire business chain — from account opening and transactions to logins and risk control.

However, the complex regulatory environments of overseas markets, combined with diverse user profiles and the rise of AIGC-driven fraud techniques, have turned identity verification into one of the most difficult challenges in global financial operations: How can institutions effectively block attacks while maintaining a smooth and seamless experience for genuine users?

The “Trade-Off” Dilemma: Balancing Detection and Approval Rates

In the past, fraudsters mainly relied on forged documents or stolen photos. Today, generative AI has dramatically lowered the cost of forgery — deepfake videos, synthetic faces, and real-time face swaps can now be produced in seconds with lifelike accuracy, rendering traditional verification methods almost defenseless.

At the same time, user diversity further increases the difficulty of accurate verification. Different countries have varying document formats, image quality standards, lighting conditions, and appearance norms. Overly strict verification rules can reject genuine users, hurting conversion rates, while overly lenient rules open the door to fraud.

FinAuth: Finding the “Golden Balance” in Dynamic Risk Control

As a financial-grade KYC identity verification platform, FinAuth integrates multi-layered anti-attack mechanisms and AI-driven configuration algorithms to defend against various types of fraud while optimizing performance for different regions.

Its goal: to continuously approach the “golden balance” — the point where security and user experience coexist at their best.

Multi-Layer Defense: Accurate Detection of AIGC-Based Attacks

FinAuth introduces an innovative “Distance Liveness Detection” technology, which randomly adjusts the recognition distance to effectively counter synthetic image and video injection attacks.

Meanwhile, the Active Defense System analyzes device characteristics, user behavior, and image attributes across multiple dimensions to proactively identify and block risky requests.

In particular, FinAuth enhances H5 environment detection in browser-based verification scenarios — significantly improving attack detection rates and overcoming a long-standing challenge in the industry.

Scenario-Based Modeling: Enhancing Pass Rates and User Experience

For different countries and user groups, FinAuth optimizes its image recognition algorithms to adapt to regional traits such as beards, head coverings, and distinctive makeup styles, improving pass rates across demographics.

Additionally, FinAuth supports a base image quality pre-check to evaluate whether an image meets the recognition threshold before matching, thereby reducing rejections caused by poor-quality images.

The system also optimizes performance under complex lighting conditions, minimizing recognition failures and boosting verification success and business efficiency.

Beyond Functionality: The Strategic Balance Between Security and Experience

Unlike conventional verification systems, FinAuth’s advantage lies not only in technological completeness, but also in its strategy-level intelligence.

Instead of simply adjusting thresholds for different scenarios, FinAuth leverages AI-based configuration algorithms to dynamically adapt verification policies according to device signals, user behavior, and detected fraud intent.

This allows FinAuth to “understand” suspicious requests rather than rely on a one-size-fits-all rule — avoiding the typical trade-off where stricter rules increase security but severely reduce pass rates.

This fine-grained operational logic is key to achieving maximum security with minimal user friction.

Case Study: The “Two-Sided Dilemma” of a Southeast Asian Digital Bank and FinAuth’s Balanced Approach

The Challenge: A Tug-of-War Between Security and Growth

In 2023, a fast-growing digital bank in Southeast Asia faced a typical growth dilemma. To combat a surge in AIGC-generated fraud attacks, its existing verification provider took the most direct approach — raising the strictness threshold of liveness detection.

While this move increased the attack detection rate to 95%, it also had a severe side effect: the pass rate for legitimate users dropped from 92% to 75%.

Many real customers — especially those using older devices or operating in low-light conditions — were mistakenly flagged as “high risk” and blocked. Customer complaints surged, and business growth stalled.

This is the typical consequence of threshold-based overcorrection — rejecting a large number of legitimate users just to stop a few fraudulent ones.

The Solution: FinAuth’s Dynamic Strategy and Precision Defense

After adopting FinAuth, the bank experienced a fundamental turnaround. FinAuth did not simply raise or lower a global threshold; instead, it applied an AI-driven dynamic configuration strategy that differentiated responses based on user risk levels.

Through specialized AIGC anti-attack algorithms and device-level security strategies, FinAuth increased the detection rate to over 99% — without sacrificing user experience or pass rates, successfully neutralizing AIGC-driven fraud operations.

Conclusion

In this ongoing battle between attack and defense, FinAuth stands out by combining security precision with user-centric experience design. It not only upholds compliance and risk control standards but also helps enterprises increase approval and conversion rates, ensuring smoother onboarding and operation for legitimate users.

As AIGC technologies continue to evolve, the confrontation between identity verification and intelligent fraud will only intensify. Like a finely balanced scale, FinAuth uses continuously advancing AI technology as its counterweight — helping global financial institutions find their “golden balance” between security and growth in an increasingly complex international environment.

January 22, 2026