Face recognition systems typically face actual challenges like facial pose, illumination, occlusion, and ageing that significantly impact the recognition accuracy. In this paper, a robust face recognition system that uses Multi-task Cascaded Convolutional Networks (MTCNN) for face detection and face alignment with an enhanced FaceNet for facial embedding extraction is presented. The enhanced FaceNet uses attention mechanisms to achieve more discriminative facial embeddings, especially in challenging scenarios. In addition, an Adaptive Feature Fusion module synthetically combines identity-specific embeddings with context information such as pose, lighting, and presence of masks, hence enhancing robustness and accuracy. Training takes place using the CelebA dataset, and the test is conducted independently on LFW and IJB-C to enable subject-disjoint evaluation. CelebA has over 200,000 faces of 10,177 individuals, LFW consists of 13,000+ faces of 5749 individuals in unconstrained conditions, and IJB-C has 31,000 faces and 117,000 video frames with extreme pose and occlusion changes. The system introduced here achieves 99.6% on CelebA, 94.2% on LFW, and 91.5% on IJB-C and outperforms baselines such as simple MTCNN-FaceNet, AFF-Net, and state-of-the-art models such as ArcFace, CosFace, and AdaCos. These findings demonstrate that the proposed framework generalizes effectively between datasets and is resilient in real-world scenarios.

The integration of artificial intelligence into education has driven growing interest in predicting student learning styles and developing recommendation systems that personalize learning pathways. While previous reviews examined these domains, most focus on pre-2023 research, overlooking recent methodological shifts. We conduct a systematic literature review of 40 studies published between 2017 and 2025, with emphasis on publications from 2023 to 2025 (70% of reviewed studies). Our analysis identifies three qualitative shifts: adoption of ensemble and deep learning methods over single classifiers, emergence of multimodal inputs including physiological signals, and evolution from isolated prediction to integrated adaptive systems. Beyond methodological synthesis, this review critically examines factors underlying observed trends and barriers to deployment. The Felder-Silverman Learning Style Model dominates research (58.3%) due to historical path dependency and instrument availability rather than demonstrated pedagogical superiority. While ensemble methods achieve high reported accuracy (87–98%), methodological concerns emerge: 65% of studies employ random rather than temporal validation, potentially inflating performance, and only 23% address production-level requirements, including privacy, scalability, and integration. We systematically analyze implementation barriers spanning computational requirements, LMS integration, educator acceptance, ethical considerations, and scalability—revealing that the gap between research prototypes and deployable systems remains substantial. Our contributions include a stakeholder impact framework, evaluation metrics taxonomy, critical analysis of reported performance claims, and identification of five research gaps with actionable recommendations. This review offers researchers and practitioners both a comprehensive synthesis of advances and a critical roadmap for bridging the implementation gap in AI-powered personalized education.