SELM: Siamese Extreme Learning Machine with Application to Face Biometrics

TL;DR

The paper addresses the challenge of face verification within Extreme Learning Machines by introducing SELM, a Siamese extension that processes two inputs in parallel, and a demographic-aware GED triplet feature to learn group-specific facial representations. The proposed five-stage framework leverages ResNet-50 features, demographic prediction, and triplet models trained on DiveFace, with verification performed in a Siamese ELM setting. Empirical results show SELM with GED features achieves state-of-the-art-like accuracy and AUC, while dramatically reducing false acceptances compared to ResNet baselines, and demonstrates clear advantages over non-Siamese architectures. The work contributes a fast, scalable, and bias-aware approach to face biometrics with strong practical implications for robust, fair face verification systems.

Abstract

Extreme Learning Machine is a powerful classification method very competitive existing classification methods. It is extremely fast at training. Nevertheless, it cannot perform face verification tasks properly because face verification tasks require comparison of facial images of two individuals at the same time and decide whether the two faces identify the same person. The structure of Extreme Leaning Machine was not designed to feed two input data streams simultaneously, thus, in 2-input scenarios Extreme Learning Machine methods are normally applied using concatenated inputs. However, this setup consumes two times more computational resources and it is not optimized for recognition tasks where learning a separable distance metric is critical. For these reasons, we propose and develop a Siamese Extreme Learning Machine (SELM). SELM was designed to be fed with two data streams in parallel simultaneously. It utilizes a dual-stream Siamese condition in the extra Siamese layer to transform the data before passing it along to the hidden layer. Moreover, we propose a Gender-Ethnicity-Dependent triplet feature exclusively trained on a variety of specific demographic groups. This feature enables learning and extracting of useful facial features of each group. Experiments were conducted to evaluate and compare the performances of SELM, Extreme Learning Machine, and DCNN. The experimental results showed that the proposed feature was able to perform correct classification at 97.87% accuracy and 99.45% AUC. They also showed that using SELM in conjunction with the proposed feature provided 98.31% accuracy and 99.72% AUC. They outperformed the well-known DCNN and Extreme Leaning Machine methods by a wide margin.

Figures (11)

• Figure 1: Siamese network concept.
• Figure 2: Weighted Similarity Extreme Learning Machine architecture.
• Figure 3: Siamese Extreme Learning Machine architecture.
• Figure 4: Triplet network structure (Net can be a DCNN).
• Figure 5: The workflow of the proposed framework.