Latest research digests
Kareem Darwish
Generative AI & Large Language Models
Large language models | Generative AI | Vision-language models | Agents | Retrieval-augmented generation | Chain-of-thought reasoning | Prompt engineering | Code generation | Text-to-image generation
Domains & Applications
Healthcare | Education | Manufacturing | Transportation | Cybersecurity | Mental health | Robotics | Autonomous systems | Autonomous driving | Dynamic environments
Learning Paradigms
Machine learning | Deep learning | Reinforcement learning | Unsupervised learning | Semi-supervised learning | Self-supervised learning | Active learning | Transfer learning | Meta-learning | Continual learning | Federated learning | Multi-task learning | Representation learning | Contrastive learning | Few-shot learning | In-context learning
Architectures & Core Models
Transformer models | Foundation models | Graph neural networks | Mixture-of-experts | Attention mechanism
Computer Vision & Graphics
Object detection | Image segmentation | Semantic segmentation | Optical character recognition | Remote sensing | Gaussian splatting | Deepfake detection
Natural Language & Speech Processing
Natural language processing | Machine translation | Speech recognition | Text-to-speech | Low-resource languages
Methodology, Tuning & Data
Anomaly detection | Fine-tuning | Parameter-efficient fine-tuning | Domain adaptation | Domain generalization | Data augmentation | Feature extraction | Uncertainty estimation | Sample efficiency | Class imbalance | Benchmarking | Time series forecasting
Safety, Efficiency & Human Factors
Ethics | Explainable AI | Interpretability | Human-AI collaboration | Robustness | Adversarial training | Differential privacy | Formal verification | Energy efficiency | Edge computing | Computational complexity | Catastrophic forgetting
Kareem Darwish
A Leap Forward for Arabic AI: From Dialects to Digital Sovereignty