Learning by Planning: Language-Guided Global Image Editing
This addresses the need for interpretable and high-resolution image editing tools for users in creative and design fields, though it is incremental as it builds on existing language-guided editing methods.
The paper tackles the problem of language-guided global image editing by developing a text-to-operation model that maps editing requests into interpretable, differentiable operations, and proposes an operation planning algorithm to generate pseudo ground truth for stable training. Results show advantages on the MA5k-Req and GIER datasets, with code made available.
Recently, language-guided global image editing draws increasing attention with growing application potentials. However, previous GAN-based methods are not only confined to domain-specific, low-resolution data but also lacking in interpretability. To overcome the collective difficulties, we develop a text-to-operation model to map the vague editing language request into a series of editing operations, e.g., change contrast, brightness, and saturation. Each operation is interpretable and differentiable. Furthermore, the only supervision in the task is the target image, which is insufficient for a stable training of sequential decisions. Hence, we propose a novel operation planning algorithm to generate possible editing sequences from the target image as pseudo ground truth. Comparison experiments on the newly collected MA5k-Req dataset and GIER dataset show the advantages of our methods. Code is available at https://jshi31.github.io/T2ONet.