Dynamic Prompt Learning: Addressing Cross-Attention Leakage for Text-Based Image Editing
Authors: kai wang, Fei Yang, Shiqi Yang, Muhammad Atif Butt, Joost van de Weijer
NeurIPS 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our method DPL, based on the publicly available Stable Diffusion, is extensively evaluated on a wide range of images, and consistently obtains superior results both quantitatively (CLIP score, Structure-Dist) and qualitatively (on user-evaluation). |
| Researcher Affiliation | Academia | 1Computer Vision Center, Barcelona, Spain 2 Universitat Autonoma de Barcelona, Barcelona, Spain 3 College of Computer Science, Nankai University, Tianjin, China |
| Pseudocode | Yes | An illustration of our method is shown in Fig. 2 and Algorithm 1. ... Algorithm 1: Dynamic Prompt Learning |
| Open Source Code | No | The paper does not provide concrete access to its own source code through a specific repository link, an explicit code release statement, or by including it in supplementary materials accessible from the main paper. It mentions external tools with links (e.g., clip retrieval), but not their own implementation code. |
| Open Datasets | Yes | We use clip retrieval [3] to obtain experimental multi-object real images from the LAION-5B [39] dataset... |
| Dataset Splits | Yes | For the ablation study, we select 100 multi-object images from various search prompts (including concepts as bird, dog, cat, book, clock, pen, pizza, table, horse, sheep, etc.). This Multi-Object set is abbreviated with MO-Set. Of the images, 40 images are used for validation and the other 60 are test set. |
| Hardware Specification | Yes | All experiments are done on an A40 GPU. |
| Software Dependencies | No | The paper mentions using "Stable Diffusion v1.4" but does not provide specific version numbers for ancillary software dependencies like programming languages (e.g., Python), deep learning frameworks (e.g., PyTorch, TensorFlow), or other libraries needed to replicate the experiment. |
| Experiment Setup | Yes | We use the validation set to optimize the various hyperparameters (see Eq. 9 and Eq. 10). ... where we set the guidance scale w = 7.5 as is standard for Stable Diffusion [17, 28, 35]. ... Set T = 50 and scale w = 1; ... Set guidance scale w = 7.5; |