Bokehlicious: Photorealistic Bokeh Rendering with Controllable Apertures

TL;DR

This work tackles the need for realistic, controllable Bokeh rendering without depth maps by introducing RealBokeh, a large real-world dataset with varying aperture and focal length, and Bokehlicious, an efficient end-to-end model with Aperture-Aware Attention that mimics lens physics. RealBokeh enables robust training and benchmarking, including a fixed-aperture subset RealBokeh_bin for fair comparisons. Bokehlicious achieves state-of-the-art results on conventional and aperture-controlled Bokeh benchmarks with significantly reduced compute and strong zero-shot generalization, and extends to defocus deblurring via RealDOF. The authors release code and data to encourage further research and practical applications in computational photography.

Abstract

Bokeh rendering methods play a key role in creating the visually appealing, softly blurred backgrounds seen in professional photography. While recent learning-based approaches show promising results, generating realistic Bokeh with variable strength remains challenging. Existing methods require additional inputs and suffer from unrealistic Bokeh reproduction due to reliance on synthetic data. In this work, we propose Bokehlicious, a highly efficient network that provides intuitive control over Bokeh strength through an Aperture-Aware Attention mechanism, mimicking the physical lens aperture. To further address the lack of high-quality real-world data, we present RealBokeh, a novel dataset featuring 23,000 high-resolution (24-MP) images captured by professional photographers, covering diverse scenes with varied aperture and focal length settings. Evaluations on both our new RealBokeh and established Bokeh rendering benchmarks show that Bokehlicious consistently outperforms SOTA methods while significantly reducing computational cost and exhibiting strong zero-shot generalization. Our method and dataset further extend to defocus deblurring, achieving competitive results on the RealDOF benchmark. Our code and data can be found at https://github.com/TimSeizinger/Bokehlicious

Figures (12)

• Figure 1: Our proposed Bokehlicious architecture trained on our new RealBokeh dataset can create highly photorealistic Bokeh effects of varying intensity, without the need for depth maps or other auxiliary inputs. Our method is able to maintain difficult foreground details like hair and excels at rendering highly complex Bokeh phenomena while maintaining low computational complexity. Some images from wadhwa2018synthetic.
• Figure 2: Visualization of PSFs. Ours mimics the optical abbreviations visible in the real PSF while the handcrafted PSF of BoMe BokehMeHybrid lacks detail and is increasingly inaccurate for large f-stops.
• Figure 3: Sample scenes from RealBokeh. Note the perfect alignment and pronounced Bokeh effect varying with the aperture setting.
• Figure 4: Overview of our propoesd Bokehlicious architecture.
• Figure 5: Each of the $N$ parallel heads within our AAA has a individual decay mask size, allowing it to attend to particular blur kernels. The masks are further tuned towards rendering specific f-stops via the signal from our AE \ref{['eq:AVEnc']}.