Safe Stable Diffusion
Safe Stable Diffusion was proposed in Safe Latent Diffusion: Mitigating Inappropriate Degeneration in Diffusion Models and mitigates inappropriate degeneration from Stable Diffusion models because they’re trained on unfiltered web-crawled datasets. For instance Stable Diffusion may unexpectedly generate nudity, violence, images depicting self-harm, and otherwise offensive content. Safe Stable Diffusion is an extension of Stable Diffusion that drastically reduces this type of content.
The abstract from the paper is:
Text-conditioned image generation models have recently achieved astonishing results in image quality and text alignment and are consequently employed in a fast-growing number of applications. Since they are highly data-driven, relying on billion-sized datasets randomly scraped from the internet, they also suffer, as we demonstrate, from degenerated and biased human behavior. In turn, they may even reinforce such biases. To help combat these undesired side effects, we present safe latent diffusion (SLD). Specifically, to measure the inappropriate degeneration due to unfiltered and imbalanced training sets, we establish a novel image generation test bed-inappropriate image prompts (I2P)-containing dedicated, real-world image-to-text prompts covering concepts such as nudity and violence. As our exhaustive empirical evaluation demonstrates, the introduced SLD removes and suppresses inappropriate image parts during the diffusion process, with no additional training required and no adverse effect on overall image quality or text alignment.
Tips
Use the safety_concept
property of StableDiffusionPipelineSafe to check and edit the current safety concept:
>>> from diffusers import StableDiffusionPipelineSafe
>>> pipeline = StableDiffusionPipelineSafe.from_pretrained("AIML-TUDA/stable-diffusion-safe")
>>> pipeline.safety_concept
'an image showing hate, harassment, violence, suffering, humiliation, harm, suicide, sexual, nudity, bodily fluids, blood, obscene gestures, illegal activity, drug use, theft, vandalism, weapons, child abuse, brutality, cruelty'
For each image generation the active concept is also contained in StableDiffusionSafePipelineOutput
.
There are 4 configurations (SafetyConfig.WEAK
, SafetyConfig.MEDIUM
, SafetyConfig.STRONG
, and SafetyConfig.MAX
) that can be applied:
>>> from diffusers import StableDiffusionPipelineSafe
>>> from diffusers.pipelines.stable_diffusion_safe import SafetyConfig
>>> pipeline = StableDiffusionPipelineSafe.from_pretrained("AIML-TUDA/stable-diffusion-safe")
>>> prompt = "the four horsewomen of the apocalypse, painting by tom of finland, gaston bussiere, craig mullins, j. c. leyendecker"
>>> out = pipeline(prompt=prompt, **SafetyConfig.MAX)
Make sure to check out the Stable Diffusion Tips section to learn how to explore the tradeoff between scheduler speed and quality, and how to reuse pipeline components efficiently!
StableDiffusionPipelineSafe
class diffusers.StableDiffusionPipelineSafe
< source >( vae: AutoencoderKL text_encoder: CLIPTextModel tokenizer: CLIPTokenizer unet: UNet2DConditionModel scheduler: KarrasDiffusionSchedulers safety_checker: SafeStableDiffusionSafetyChecker feature_extractor: CLIPImageProcessor image_encoder: Optional = None requires_safety_checker: bool = True )
Parameters
- vae (AutoencoderKL) — Variational Auto-Encoder (VAE) model to encode and decode images to and from latent representations.
- text_encoder (CLIPTextModel) — Frozen text-encoder (clip-vit-large-patch14).
- tokenizer (CLIPTokenizer) —
A
CLIPTokenizer
to tokenize text. - unet (UNet2DConditionModel) —
A
UNet2DConditionModel
to denoise the encoded image latents. - scheduler (SchedulerMixin) —
A scheduler to be used in combination with
unet
to denoise the encoded image latents. Can be one of DDIMScheduler, LMSDiscreteScheduler, or PNDMScheduler. - safety_checker (
StableDiffusionSafetyChecker
) — Classification module that estimates whether generated images could be considered offensive or harmful. Please refer to the model card for more details about a model’s potential harms. - feature_extractor (CLIPImageProcessor) —
A
CLIPImageProcessor
to extract features from generated images; used as inputs to thesafety_checker
.
Pipeline based on the StableDiffusionPipeline for text-to-image generation using Safe Latent Diffusion.
This model inherits from DiffusionPipeline. Check the superclass documentation for the generic methods implemented for all pipelines (downloading, saving, running on a particular device, etc.).
The pipeline also inherits the following loading methods:
- load_ip_adapter() for loading IP Adapters
__call__
< source >( prompt: Union height: Optional = None width: Optional = None num_inference_steps: int = 50 guidance_scale: float = 7.5 negative_prompt: Union = None num_images_per_prompt: Optional = 1 eta: float = 0.0 generator: Union = None latents: Optional = None ip_adapter_image: Union = None output_type: Optional = 'pil' return_dict: bool = True callback: Optional = None callback_steps: int = 1 sld_guidance_scale: Optional = 1000 sld_warmup_steps: Optional = 10 sld_threshold: Optional = 0.01 sld_momentum_scale: Optional = 0.3 sld_mom_beta: Optional = 0.4 ) → StableDiffusionPipelineOutput or tuple
Parameters
- prompt (
str
orList[str]
) — The prompt or prompts to guide image generation. If not defined, you need to passprompt_embeds
. - height (
int
, optional, defaults toself.unet.config.sample_size * self.vae_scale_factor
) — The height in pixels of the generated image. - width (
int
, optional, defaults toself.unet.config.sample_size * self.vae_scale_factor
) — The width in pixels of the generated image. - num_inference_steps (
int
, optional, defaults to 50) — The number of denoising steps. More denoising steps usually lead to a higher quality image at the expense of slower inference. - guidance_scale (
float
, optional, defaults to 7.5) — A higher guidance scale value encourages the model to generate images closely linked to the textprompt
at the expense of lower image quality. Guidance scale is enabled whenguidance_scale > 1
. - negative_prompt (
str
orList[str]
, optional) — The prompt or prompts to guide what to not include in image generation. If not defined, you need to passnegative_prompt_embeds
instead. Ignored when not using guidance (guidance_scale < 1
). - num_images_per_prompt (
int
, optional, defaults to 1) — The number of images to generate per prompt. - eta (
float
, optional, defaults to 0.0) — Corresponds to parameter eta (η) from the DDIM paper. Only applies to the DDIMScheduler, and is ignored in other schedulers. - generator (
torch.Generator
orList[torch.Generator]
, optional) — Atorch.Generator
to make generation deterministic. - latents (
torch.Tensor
, optional) — Pre-generated noisy latents sampled from a Gaussian distribution, to be used as inputs for image generation. Can be used to tweak the same generation with different prompts. If not provided, a latents tensor is generated by sampling using the supplied randomgenerator
. ip_adapter_image — (PipelineImageInput
, optional): Optional image input to work with IP Adapters. - output_type (
str
, optional, defaults to"pil"
) — The output format of the generated image. Choose betweenPIL.Image
ornp.array
. - return_dict (
bool
, optional, defaults toTrue
) — Whether or not to return a StableDiffusionPipelineOutput instead of a plain tuple. - callback (
Callable
, optional) — A function that calls everycallback_steps
steps during inference. The function is called with the following arguments:callback(step: int, timestep: int, latents: torch.Tensor)
. - callback_steps (
int
, optional, defaults to 1) — The frequency at which thecallback
function is called. If not specified, the callback is called at every step. - sld_guidance_scale (
float
, optional, defaults to 1000) — Ifsld_guidance_scale < 1
, safety guidance is disabled. - sld_warmup_steps (
int
, optional, defaults to 10) — Number of warmup steps for safety guidance. SLD is only be applied for diffusion steps greater thansld_warmup_steps
. - sld_threshold (
float
, optional, defaults to 0.01) — Threshold that separates the hyperplane between appropriate and inappropriate images. - sld_momentum_scale (
float
, optional, defaults to 0.3) — Scale of the SLD momentum to be added to the safety guidance at each diffusion step. If set to 0.0, momentum is disabled. Momentum is built up during warmup for diffusion steps smaller thansld_warmup_steps
. - sld_mom_beta (
float
, optional, defaults to 0.4) — Defines how safety guidance momentum builds up.sld_mom_beta
indicates how much of the previous momentum is kept. Momentum is built up during warmup for diffusion steps smaller thansld_warmup_steps
.
Returns
StableDiffusionPipelineOutput or tuple
If return_dict
is True
, StableDiffusionPipelineOutput is returned,
otherwise a tuple
is returned where the first element is a list with the generated images and the
second element is a list of bool
s indicating whether the corresponding generated image contains
“not-safe-for-work” (nsfw) content.
The call function to the pipeline for generation.
Examples:
import torch
from diffusers import StableDiffusionPipelineSafe
from diffusers.pipelines.stable_diffusion_safe import SafetyConfig
pipeline = StableDiffusionPipelineSafe.from_pretrained(
"AIML-TUDA/stable-diffusion-safe", torch_dtype=torch.float16
).to("cuda")
prompt = "the four horsewomen of the apocalypse, painting by tom of finland, gaston bussiere, craig mullins, j. c. leyendecker"
image = pipeline(prompt=prompt, **SafetyConfig.MEDIUM).images[0]
StableDiffusionSafePipelineOutput
class diffusers.pipelines.stable_diffusion_safe.StableDiffusionSafePipelineOutput
< source >( images: Union nsfw_content_detected: Optional unsafe_images: Union applied_safety_concept: Optional )
Parameters
- images (
List[PIL.Image.Image]
ornp.ndarray
) — List of denoised PIL images of lengthbatch_size
or numpy array of shape(batch_size, height, width, num_channels)
. PIL images or numpy array present the denoised images of the diffusion pipeline. - nsfw_content_detected (
List[bool]
) — List of flags denoting whether the corresponding generated image likely represents “not-safe-for-work” (nsfw) content, orNone
if safety checking could not be performed. - images (
List[PIL.Image.Image]
ornp.ndarray
) — List of denoised PIL images that were flagged by the safety checker any may contain “not-safe-for-work” (nsfw) content, orNone
if no safety check was performed or no images were flagged. - applied_safety_concept (
str
) — The safety concept that was applied for safety guidance, orNone
if safety guidance was disabled
Output class for Safe Stable Diffusion pipelines.