From background to Blending problem

[AzureRossi]

First, thanks for the great work but I may have the following problem. I downloaded the background image you provided and configured it with the same parameters as follows: old man sitting, high quality Negative prompt: bad, ugly Steps: 20, Sampler: DPM++ 2M SDE Karras, CFG scale: 7, Seed: 12345, Size: 896x1152, Model hash: 1fe6c7ec54, Model: juggernautXL_version6Rundiffusion, layerdiffusion_enabled: True, layerdiffusion_method: From Background to Blending, layerdiffusion_weight: 1, layerdiffusion_ending_step: 1, layerdiffusion_fg_image: False, layerdiffusion_bg_image: True, layerdiffusion_blend_image: False, layerdiffusion_resize_mode: Crop and Resize, Version: f0.0.17v1.8.0rc-latest-276-g29be1da7

However, my inference results are as follows:

Also, there are some other failed cases:

(1)

parameters: handlesome man sitting on the bench, high quality Negative prompt: bad, ugly Steps: 20, Sampler: DPM++ 2M SDE Karras, CFG scale: 7, Seed: 12345, Size: 896x1152, Model hash: 1fe6c7ec54, Model: juggernautXL_version6Rundiffusion, layerdiffusion_enabled: True, layerdiffusion_method: From Background to Blending, layerdiffusion_weight: 1, layerdiffusion_ending_step: 1, layerdiffusion_fg_image: False, layerdiffusion_bg_image: True, layerdiffusion_blend_image: False, layerdiffusion_resize_mode: Crop and Resize, Version: f0.0.17v1.8.0rc-latest-276-g29be1da7

(2)

input background: parameters: plants on table, high quality Negative prompt: bad, ugly Steps: 20, Sampler: DPM++ 2M SDE Karras, CFG scale: 7, Seed: 12345, Size: 896x1152, Model hash: 1fe6c7ec54, Model: juggernautXL_version6Rundiffusion, layerdiffusion_enabled: True, layerdiffusion_method: From Background to Blending, layerdiffusion_weight: 1, layerdiffusion_ending_step: 1, layerdiffusion_fg_image: False, layerdiffusion_bg_image: True, layerdiffusion_blend_image: False, layerdiffusion_resize_mode: Crop and Resize, Version: f0.0.17v1.8.0rc-latest-276-g29be1da7

The result always involves the generation of new objects in certain areas, while the geometric structures between layers are not well aligned. I would like to know if this is a problem with layer_xl_bg2ble.safetensors or if it's due to some details I might have overlooked personally. Thanks.

[DuoLi1999]

Yeah, am also meeting these situations, have you fix it? Or this is the normal performance?

[AzureRossi]

Yeah, am also meeting these situations, have you fix it? Or this is the normal performance?

Sorry for the late reply. I think the possible issue could be the selection of the seed. Sometimes the foreground might come with some redundant elements (probably related to its original SD) as follow:

In the later released version 1.5 of the SD, the author also mentioned the possibility of injecting prompts individually for each layer to solve similar issues (the author mentioned the issue of removing extra apples from the background). I think it might be resolved by: 1. Prompting individual layers for foreground 2. Selecting seeds without issues to address the above problem. Of course, feel free to propose any new ideas or solutions you might have recently.