DiffBMP: Differentiable Rendering with Bitmap Primitives

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hong, Seongmin, Kim, Junghun James, Kim, Daehyeop, Chung, Insoo, Chun, Se Young
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908908250988544
author Hong, Seongmin
Kim, Junghun James
Kim, Daehyeop
Chung, Insoo
Chun, Se Young
author_facet Hong, Seongmin
Kim, Junghun James
Kim, Daehyeop
Chung, Insoo
Chun, Se Young
contents We introduce DiffBMP, a scalable and efficient differentiable rendering engine for a collection of bitmap images. Our work addresses a limitation that traditional differentiable renderers are constrained to vector graphics, given that most images in the world are bitmaps. Our core contribution is a highly parallelized rendering pipeline, featuring a custom CUDA implementation for calculating gradients. This system can, for example, optimize the position, rotation, scale, color, and opacity of thousands of bitmap primitives all in under 1 min using a consumer GPU. We employ and validate several techniques to facilitate the optimization: soft rasterization via Gaussian blur, structure-aware initialization, noisy canvas, and specialized losses/heuristics for videos or spatially constrained images. We demonstrate DiffBMP is not just an isolated tool, but a practical one designed to integrate into creative workflows. It supports exporting compositions to a native, layered file format, and the entire framework is publicly accessible via an easy-to-hack Python package.
format Preprint
id arxiv_https___arxiv_org_abs_2602_22625
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle DiffBMP: Differentiable Rendering with Bitmap Primitives
Hong, Seongmin
Kim, Junghun James
Kim, Daehyeop
Chung, Insoo
Chun, Se Young
Graphics
Computer Vision and Pattern Recognition
We introduce DiffBMP, a scalable and efficient differentiable rendering engine for a collection of bitmap images. Our work addresses a limitation that traditional differentiable renderers are constrained to vector graphics, given that most images in the world are bitmaps. Our core contribution is a highly parallelized rendering pipeline, featuring a custom CUDA implementation for calculating gradients. This system can, for example, optimize the position, rotation, scale, color, and opacity of thousands of bitmap primitives all in under 1 min using a consumer GPU. We employ and validate several techniques to facilitate the optimization: soft rasterization via Gaussian blur, structure-aware initialization, noisy canvas, and specialized losses/heuristics for videos or spatially constrained images. We demonstrate DiffBMP is not just an isolated tool, but a practical one designed to integrate into creative workflows. It supports exporting compositions to a native, layered file format, and the entire framework is publicly accessible via an easy-to-hack Python package.
title DiffBMP: Differentiable Rendering with Bitmap Primitives
topic Graphics
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2602.22625