Img2Real

Img2Real is for making images look realistic; use an image as the input and aim for a realistic-looking image as the output.

Suggestion · not verified

The description lists a base model, ControlNet, face bounding-box detection, eye segmentation, SAM, upscaling, and tagging.

The description does not specify an input file format or resolution.

Img2Real is listed under Workflows, is published by V01, and uses Pony as its base model.

A Pipe version does not apply to this workflow.

The workflow lists these node packs: ComfyUI_ADV_CLIP_emb, Comfyui_joytag, comfyui-impact-pack, and comfyui-impact-subpack.

It also lists ComfyUI-string-util, comfyui-wd14-tagger, derfuu_comfyui_moddednodes, and comfyui_controlnet_aux.

Other listed files are bbox/face_yolov8m.pt, diffusion_pytorch_model_promax.safetensors, and tamePonyThe_v25.safetensors.

For the listed layout, place TAME Pony in ComfyUI\models\checkpoints and ControlNet Union SDXL ProMax 1.0 in ComfyUI\models\controlnet\xinsir.

Suggestion · not verified

Place face_yolov8m in ComfyUI\models\ultralytics\bbox, Anzhc Eyes -seg-hd in ComfyUI\models\ultralytics\segm, and sam_vit_h_4b8939 in ComfyUI\models\sams.

Suggestion · not verified

Place 4x-UltraSharp in ComfyUI\models\upscale_models and JoyTag in ComfyUI\custom_nodes\Comfyui_joytag\checkpoints.

Suggestion · not verified

Before setup, verify that bbox/face_yolov8m.pt, tamePonyThe_v25.safetensors, and diffusion_pytorch_model_promax.safetensors are available.

Suggestion · not verified

Before setup, confirm the exact node-pack installations for comfyui_controlnet_aux and was-node-suite-comfyui.

Suggestion · not verified