Realistic indoor and outdoor image synthesis remains a core challenge in computer vision and graphics. While learning-based methods offer ease of use, they often lack physical consistency. On the other hand, traditional physics-based rendering (PBR) techniques can produce highly realistic images but are computationally expensive. Intrinsic image representation provides a promising middle ground, breaking down images into their fundamental components—such as geometry, material, and lighting—to enable controlled synthesis. However, conventional PBR materials struggle with complex surface models, particularly those involving high specular reflections and transparent surfaces. In this work, researchers extend intrinsic image representation to include reflection and transmission properties, thereby improving the synthesis of transparent materials like glass and windows. They propose an explicit intrinsic synthesis framework that ensures deterministic and interpretable image generation. By incorporating extended PBR materials (ePBR materials), this framework allows for effective material editing and precise control over the synthesized images. This advancement addresses a significant limitation of existing PBR methods, making it possible to create more realistic and versatile images. The ability to accurately model and manipulate transparent and reflective surfaces opens up new possibilities for applications in virtual reality, augmented reality, and advanced imaging systems. The research underscores the importance of combining machine learning techniques with physical principles to achieve both efficiency and realism in image synthesis.