Ef-QuantFace: Streamlined Face Recognition with Small Data and Low-Bit Precision
This reduces data and training time requirements for face recognition model compression, though it is incremental in the field of model quantization.
The paper tackles the problem of data-intensive model quantization for face recognition by fine-tuning with only up to 14,000 images, 440 times smaller than traditional datasets, and achieves 96.15% accuracy on the IJB-C dataset.
In recent years, model quantization for face recognition has gained prominence. Traditionally, compressing models involved vast datasets like the 5.8 million-image MS1M dataset as well as extensive training times, raising the question of whether such data enormity is essential. This paper addresses this by introducing an efficiency-driven approach, fine-tuning the model with just up to 14,000 images, 440 times smaller than MS1M. We demonstrate that effective quantization is achievable with a smaller dataset, presenting a new paradigm. Moreover, we incorporate an evaluation-based metric loss and achieve an outstanding 96.15% accuracy on the IJB-C dataset, establishing a new state-of-the-art compressed model training for face recognition. The subsequent analysis delves into potential applications, emphasizing the transformative power of this approach. This paper advances model quantization by highlighting the efficiency and optimal results with small data and training time.