import json
import os
import uuid
import cv2
import gradio as gr
import numpy as np
import spaces
import torch
import torchvision
from diffusers import AutoencoderKL, DDIMScheduler
from einops import rearrange
from huggingface_hub import hf_hub_download
from omegaconf import OmegaConf
from PIL import Image
from torchvision import transforms
from transformers import CLIPTextModel, CLIPTokenizer
from modules.unet import UNet3DConditionFlowModel
from pipelines.pipeline_imagecoductor import ImageConductorPipeline
from utils.gradio_utils import ensure_dirname, image2pil, split_filename, visualize_drag
from utils.lora_utils import add_LoRA_to_controlnet
from utils.utils import (
bivariate_Gaussian,
create_flow_controlnet,
create_image_controlnet,
interpolate_trajectory,
load_model,
load_weights,
)
from utils.visualizer import vis_flow_to_video
#### Description ####
title = r"""CustomNet: Object Customization with Variable-Viewpoints in Text-to-Image Diffusion Models
"""
head = r"""
Image Conductor: Precision Control for Interactive Video Synthesis
"""
descriptions = r"""
Official Gradio Demo for Image Conductor: Precision Control for Interactive Video Synthesis.
🧙Image Conductor enables precise, fine-grained control for generating motion-controllable videos from images, advancing the practical application of interactive video synthesis.
"""
instructions = r"""
- ⭐️ step1: Upload or select one image from Example.
- ⭐️ step2: Click 'Add Drag' to draw some drags.
- ⭐️ step3: Input text prompt that complements the image (Necessary).
- ⭐️ step4: Select 'Drag Mode' to specify the control of camera transition or object movement.
- ⭐️ step5: Click 'Run' button to generate video assets.
- ⭐️ others: Click 'Delete last drag' to delete the whole lastest path. Click 'Delete last step' to delete the lastest clicked control point.
"""
citation = r"""
If Image Conductor is helpful, please help to ⭐ the Github Repo. Thanks!
[](https://github.com/liyaowei-stu/ImageConductor)
---
📝 **Citation**
If our work is useful for your research, please consider citing:
```bibtex
@misc{li2024imageconductor,
title={Image Conductor: Precision Control for Interactive Video Synthesis},
author={Li, Yaowei and Wang, Xintao and Zhang, Zhaoyang and Wang, Zhouxia and Yuan, Ziyang and Xie, Liangbin and Zou, Yuexian and Shan, Ying},
year={2024},
eprint={2406.15339},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
📧 **Contact**
If you have any questions, please feel free to reach me out at ywl@stu.pku.edu.cn.
# """
flow_controlnet_path = hf_hub_download("TencentARC/ImageConductor", "flow_controlnet.ckpt")
image_controlnet_path = hf_hub_download("TencentARC/ImageConductor", "image_controlnet.ckpt")
unet_path = hf_hub_download("TencentARC/ImageConductor", "unet.ckpt")
helloobjects_path = hf_hub_download("TencentARC/ImageConductor", "helloobjects_V12c.safetensors")
tusun_path = hf_hub_download("TencentARC/ImageConductor", "TUSUN.safetensors")
os.makedirs("models/sd1-5", exist_ok=True)
sd15_config_path = hf_hub_download("runwayml/stable-diffusion-v1-5", "config.json", subfolder="unet")
if not os.path.exists("models/sd1-5/config.json"):
os.symlink(sd15_config_path, "models/sd1-5/config.json")
if not os.path.exists("models/sd1-5/unet.ckpt"):
os.symlink(unet_path, "models/sd1-5/unet.ckpt")
# mv1 = os.system(f'mv /usr/local/lib/python3.10/site-packages/gradio/helpers.py /usr/local/lib/python3.10/site-packages/gradio/helpers_bkp.py')
# mv2 = os.system(f'mv helpers.py /usr/local/lib/python3.10/site-packages/gradio/helpers.py')
# # 检查命令是否成功
# if mv1 == 0 and mv2 == 0:
# print("file move success!")
# else:
# print("file move failed!")
# - - - - - examples - - - - - #
image_examples = [
[
"__asset__/images/object/turtle-1.jpg",
"a sea turtle gracefully swimming over a coral reef in the clear blue ocean.",
"object",
11318446767408804497,
"",
"turtle",
"__asset__/turtle.mp4",
],
[
"__asset__/images/object/rose-1.jpg",
"a red rose engulfed in flames.",
"object",
6854275249656120509,
"",
"rose",
"__asset__/rose.mp4",
],
[
"__asset__/images/object/jellyfish-1.jpg",
"intricate detailing,photorealism,hyperrealistic, glowing jellyfish mushroom, flying, starry sky, bokeh, golden ratio composition.",
"object",
17966188172968903484,
"HelloObject",
"jellyfish",
"__asset__/jellyfish.mp4",
],
[
"__asset__/images/camera/lush-1.jpg",
"detailed craftsmanship, photorealism, hyperrealistic, roaring waterfall, misty spray, lush greenery, vibrant rainbow, golden ratio composition.",
"camera",
7970487946960948963,
"HelloObject",
"lush",
"__asset__/lush.mp4",
],
[
"__asset__/images/camera/tusun-1.jpg",
"tusuncub with its mouth open, blurry, open mouth, fangs, photo background, looking at viewer, tongue, full body, solo, cute and lovely, Beautiful and realistic eye details, perfect anatomy, Nonsense, pure background, Centered-Shot, realistic photo, photograph, 4k, hyper detailed, DSLR, 24 Megapixels, 8mm Lens, Full Frame, film grain, Global Illumination, studio Lighting, Award Winning Photography, diffuse reflection, ray tracing.",
"camera",
996953226890228361,
"TUSUN",
"tusun",
"__asset__/tusun.mp4",
],
[
"__asset__/images/camera/painting-1.jpg",
"A oil painting.",
"camera",
16867854766769816385,
"",
"painting",
"__asset__/painting.mp4",
],
]
POINTS = {
"turtle": "__asset__/trajs/object/turtle-1.json",
"rose": "__asset__/trajs/object/rose-1.json",
"jellyfish": "__asset__/trajs/object/jellyfish-1.json",
"lush": "__asset__/trajs/camera/lush-1.json",
"tusun": "__asset__/trajs/camera/tusun-1.json",
"painting": "__asset__/trajs/camera/painting-1.json",
}
IMAGE_PATH = {
"turtle": "__asset__/images/object/turtle-1.jpg",
"rose": "__asset__/images/object/rose-1.jpg",
"jellyfish": "__asset__/images/object/jellyfish-1.jpg",
"lush": "__asset__/images/camera/lush-1.jpg",
"tusun": "__asset__/images/camera/tusun-1.jpg",
"painting": "__asset__/images/camera/painting-1.jpg",
}
DREAM_BOOTH = {
"HelloObject": helloobjects_path,
}
LORA = {
"TUSUN": tusun_path,
}
LORA_ALPHA = {
"TUSUN": 0.6,
}
NPROMPT = {
"HelloObject": "FastNegativeV2,(bad-artist:1),(worst quality, low quality:1.4),(bad_prompt_version2:0.8),bad-hands-5,lowres,bad anatomy,bad hands,((text)),(watermark),error,missing fingers,extra digit,fewer digits,cropped,worst quality,low quality,normal quality,((username)),blurry,(extra limbs),bad-artist-anime,badhandv4,EasyNegative,ng_deepnegative_v1_75t,verybadimagenegative_v1.3,BadDream,(three hands:1.6),(three legs:1.2),(more than two hands:1.4),(more than two legs,:1.2)"
}
output_dir = "outputs"
ensure_dirname(output_dir)
def points_to_flows(track_points, model_length, height, width):
input_drag = np.zeros((model_length - 1, height, width, 2))
for splited_track in track_points:
if len(splited_track) == 1: # stationary point
displacement_point = tuple([splited_track[0][0] + 1, splited_track[0][1] + 1])
splited_track = tuple([splited_track[0], displacement_point])
# interpolate the track
splited_track = interpolate_trajectory(splited_track, model_length)
splited_track = splited_track[:model_length]
if len(splited_track) < model_length:
splited_track = splited_track + [splited_track[-1]] * (model_length - len(splited_track))
for i in range(model_length - 1):
start_point = splited_track[i]
end_point = splited_track[i + 1]
input_drag[i][int(start_point[1])][int(start_point[0])][0] = end_point[0] - start_point[0]
input_drag[i][int(start_point[1])][int(start_point[0])][1] = end_point[1] - start_point[1]
return input_drag
class ImageConductor:
def __init__(
self, device, unet_path, image_controlnet_path, flow_controlnet_path, height, width, model_length, lora_rank=64
):
self.device = device
tokenizer = CLIPTokenizer.from_pretrained("runwayml/stable-diffusion-v1-5", subfolder="tokenizer")
text_encoder = CLIPTextModel.from_pretrained("runwayml/stable-diffusion-v1-5", subfolder="text_encoder").to(
device
)
vae = AutoencoderKL.from_pretrained("runwayml/stable-diffusion-v1-5", subfolder="vae").to(device)
inference_config = OmegaConf.load("configs/inference/inference.yaml")
unet = UNet3DConditionFlowModel.from_pretrained_2d(
"models/sd1-5/", unet_additional_kwargs=OmegaConf.to_container(inference_config.unet_additional_kwargs)
)
self.vae = vae
### >>> Initialize UNet module >>> ###
load_model(unet, unet_path)
### >>> Initialize image controlnet module >>> ###
image_controlnet = create_image_controlnet("configs/inference/image_condition.yaml", unet)
load_model(image_controlnet, image_controlnet_path)
### >>> Initialize flow controlnet module >>> ###
flow_controlnet = create_flow_controlnet("configs/inference/flow_condition.yaml", unet)
add_LoRA_to_controlnet(lora_rank, flow_controlnet)
load_model(flow_controlnet, flow_controlnet_path)
unet.eval().to(device)
image_controlnet.eval().to(device)
flow_controlnet.eval().to(device)
self.pipeline = ImageConductorPipeline(
unet=unet,
vae=vae,
tokenizer=tokenizer,
text_encoder=text_encoder,
scheduler=DDIMScheduler(**OmegaConf.to_container(inference_config.noise_scheduler_kwargs)),
image_controlnet=image_controlnet,
flow_controlnet=flow_controlnet,
).to(device)
self.height = height
self.width = width
# _, model_step, _ = split_filename(model_path)
# self.ouput_prefix = f'{model_step}_{width}X{height}'
self.model_length = model_length
blur_kernel = bivariate_Gaussian(kernel_size=99, sig_x=10, sig_y=10, theta=0, grid=None, isotropic=True)
self.blur_kernel = blur_kernel
@spaces.GPU(duration=180)
def run(
self,
first_frame_path,
tracking_points,
prompt,
drag_mode,
negative_prompt,
seed,
randomize_seed,
guidance_scale,
num_inference_steps,
personalized,
examples_type,
):
print("Run!")
if examples_type != "":
### for adapting high version gradio
tracking_points = gr.State([])
first_frame_path = IMAGE_PATH[examples_type]
points = json.load(open(POINTS[examples_type]))
tracking_points.value.extend(points)
print("example first_frame_path", first_frame_path)
print("example tracking_points", tracking_points.value)
original_width, original_height = 384, 256
if isinstance(tracking_points, list):
input_all_points = tracking_points
else:
input_all_points = tracking_points.value
print("input_all_points", input_all_points)
resized_all_points = [
tuple(
[
tuple([float(e1[0] * self.width / original_width), float(e1[1] * self.height / original_height)])
for e1 in e
]
)
for e in input_all_points
]
dir, base, ext = split_filename(first_frame_path)
id = base.split("_")[-1]
visualized_drag, _ = visualize_drag(
first_frame_path, resized_all_points, drag_mode, self.width, self.height, self.model_length
)
## image condition
image_transforms = transforms.Compose(
[
transforms.RandomResizedCrop(
(self.height, self.width), (1.0, 1.0), ratio=(self.width / self.height, self.width / self.height)
),
transforms.ToTensor(),
]
)
image_paths = [first_frame_path]
controlnet_images = [(image_transforms(Image.open(path).convert("RGB"))) for path in image_paths]
controlnet_images = torch.stack(controlnet_images).unsqueeze(0).to(device)
controlnet_images = rearrange(controlnet_images, "b f c h w -> b c f h w")
num_controlnet_images = controlnet_images.shape[2]
controlnet_images = rearrange(controlnet_images, "b c f h w -> (b f) c h w")
self.vae.to(device)
controlnet_images = self.vae.encode(controlnet_images * 2.0 - 1.0).latent_dist.sample() * 0.18215
controlnet_images = rearrange(controlnet_images, "(b f) c h w -> b c f h w", f=num_controlnet_images)
# flow condition
controlnet_flows = points_to_flows(resized_all_points, self.model_length, self.height, self.width)
for i in range(0, self.model_length - 1):
controlnet_flows[i] = cv2.filter2D(controlnet_flows[i], -1, self.blur_kernel)
controlnet_flows = np.concatenate(
[np.zeros_like(controlnet_flows[0])[np.newaxis, ...], controlnet_flows], axis=0
) # pad the first frame with zero flow
os.makedirs(os.path.join(output_dir, "control_flows"), exist_ok=True)
trajs_video = vis_flow_to_video(controlnet_flows, num_frames=self.model_length) # T-1 x H x W x 3
torchvision.io.write_video(
f"{output_dir}/control_flows/sample-{id}-train_flow.mp4",
trajs_video,
fps=8,
video_codec="h264",
options={"crf": "10"},
)
controlnet_flows = torch.from_numpy(controlnet_flows)[None][:, : self.model_length, ...]
controlnet_flows = rearrange(controlnet_flows, "b f h w c-> b c f h w").float().to(device)
dreambooth_model_path = DREAM_BOOTH.get(personalized, "")
lora_model_path = LORA.get(personalized, "")
lora_alpha = LORA_ALPHA.get(personalized, 0.6)
self.pipeline = load_weights(
self.pipeline,
dreambooth_model_path=dreambooth_model_path,
lora_model_path=lora_model_path,
lora_alpha=lora_alpha,
).to(device)
if NPROMPT.get(personalized, "") != "":
negative_prompt = NPROMPT.get(personalized)
if randomize_seed:
random_seed = torch.seed()
else:
seed = int(seed)
random_seed = seed
torch.manual_seed(random_seed)
torch.cuda.manual_seed_all(random_seed)
print(f"current seed: {torch.initial_seed()}")
sample = self.pipeline(
prompt,
negative_prompt=negative_prompt,
num_inference_steps=num_inference_steps,
guidance_scale=guidance_scale,
width=self.width,
height=self.height,
video_length=self.model_length,
controlnet_images=controlnet_images, # 1 4 1 32 48
controlnet_image_index=[0],
controlnet_flows=controlnet_flows, # [1, 2, 16, 256, 384]
control_mode=drag_mode,
eval_mode=True,
).videos
outputs_path = os.path.join(output_dir, f"output_{i}_{id}.mp4")
vis_video = (rearrange(sample[0], "c t h w -> t h w c") * 255.0).clip(0, 255)
torchvision.io.write_video(outputs_path, vis_video, fps=8, video_codec="h264", options={"crf": "10"})
# outputs_path = os.path.join(output_dir, f'output_{i}_{id}.gif')
# save_videos_grid(sample[0][None], outputs_path)
print("Done!")
return {output_image: visualized_drag, output_video: outputs_path}
def reset_states(first_frame_path, tracking_points):
first_frame_path = None
tracking_points = []
return {input_image: None, first_frame_path_var: first_frame_path, tracking_points_var: tracking_points}
def preprocess_image(image, tracking_points):
image_pil = image2pil(image.name)
raw_w, raw_h = image_pil.size
resize_ratio = max(384 / raw_w, 256 / raw_h)
image_pil = image_pil.resize((int(raw_w * resize_ratio), int(raw_h * resize_ratio)), Image.BILINEAR)
image_pil = transforms.CenterCrop((256, 384))(image_pil.convert("RGB"))
id = str(uuid.uuid4())[:4]
first_frame_path = os.path.join(output_dir, f"first_frame_{id}.jpg")
image_pil.save(first_frame_path, quality=95)
tracking_points = []
return {
input_image: first_frame_path,
first_frame_path_var: first_frame_path,
tracking_points_var: tracking_points,
personalized: "",
}
def add_tracking_points(
tracking_points, first_frame_path, drag_mode, evt: gr.SelectData
): # SelectData is a subclass of EventData
if drag_mode == "object":
color = (255, 0, 0, 255)
elif drag_mode == "camera":
color = (0, 0, 255, 255)
print(f"You selected {evt.value} at {evt.index} from {evt.target}")
if not tracking_points:
tracking_points = [[]]
tracking_points[-1].append(evt.index)
transparent_background = Image.open(first_frame_path).convert("RGBA")
w, h = transparent_background.size
transparent_layer = np.zeros((h, w, 4))
for track in tracking_points:
if len(track) > 1:
for i in range(len(track) - 1):
start_point = track[i]
end_point = track[i + 1]
vx = end_point[0] - start_point[0]
vy = end_point[1] - start_point[1]
arrow_length = np.sqrt(vx**2 + vy**2)
if i == len(track) - 2:
cv2.arrowedLine(
transparent_layer, tuple(start_point), tuple(end_point), color, 2, tipLength=8 / arrow_length
)
else:
cv2.line(
transparent_layer,
tuple(start_point),
tuple(end_point),
color,
2,
)
else:
cv2.circle(transparent_layer, tuple(track[0]), 5, color, -1)
transparent_layer = Image.fromarray(transparent_layer.astype(np.uint8))
trajectory_map = Image.alpha_composite(transparent_background, transparent_layer)
return {tracking_points_var: tracking_points, input_image: trajectory_map}
def add_drag(tracking_points):
if not tracking_points or tracking_points[-1]:
tracking_points.append([])
return {tracking_points_var: tracking_points}
def delete_last_drag(tracking_points, first_frame_path, drag_mode):
if drag_mode == "object":
color = (255, 0, 0, 255)
elif drag_mode == "camera":
color = (0, 0, 255, 255)
if tracking_points:
tracking_points.pop()
transparent_background = Image.open(first_frame_path).convert("RGBA")
w, h = transparent_background.size
transparent_layer = np.zeros((h, w, 4))
for track in tracking_points:
if len(track) > 1:
for i in range(len(track) - 1):
start_point = track[i]
end_point = track[i + 1]
vx = end_point[0] - start_point[0]
vy = end_point[1] - start_point[1]
arrow_length = np.sqrt(vx**2 + vy**2)
if i == len(track) - 2:
cv2.arrowedLine(
transparent_layer, tuple(start_point), tuple(end_point), color, 2, tipLength=8 / arrow_length
)
else:
cv2.line(
transparent_layer,
tuple(start_point),
tuple(end_point),
color,
2,
)
else:
cv2.circle(transparent_layer, tuple(track[0]), 5, color, -1)
transparent_layer = Image.fromarray(transparent_layer.astype(np.uint8))
trajectory_map = Image.alpha_composite(transparent_background, transparent_layer)
return {tracking_points_var: tracking_points, input_image: trajectory_map}
def delete_last_step(tracking_points, first_frame_path, drag_mode):
if drag_mode == "object":
color = (255, 0, 0, 255)
elif drag_mode == "camera":
color = (0, 0, 255, 255)
tracking_points[-1].pop()
transparent_background = Image.open(first_frame_path).convert("RGBA")
w, h = transparent_background.size
transparent_layer = np.zeros((h, w, 4))
for track in tracking_points:
if len(track) > 1:
for i in range(len(track) - 1):
start_point = track[i]
end_point = track[i + 1]
vx = end_point[0] - start_point[0]
vy = end_point[1] - start_point[1]
arrow_length = np.sqrt(vx**2 + vy**2)
if i == len(track) - 2:
cv2.arrowedLine(
transparent_layer, tuple(start_point), tuple(end_point), color, 2, tipLength=8 / arrow_length
)
else:
cv2.line(
transparent_layer,
tuple(start_point),
tuple(end_point),
color,
2,
)
else:
cv2.circle(transparent_layer, tuple(track[0]), 5, color, -1)
transparent_layer = Image.fromarray(transparent_layer.astype(np.uint8))
trajectory_map = Image.alpha_composite(transparent_background, transparent_layer)
return {tracking_points_var: tracking_points, input_image: trajectory_map}
device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
ImageConductor_net = ImageConductor(
device=device,
unet_path=unet_path,
image_controlnet_path=image_controlnet_path,
flow_controlnet_path=flow_controlnet_path,
height=256,
width=384,
model_length=16,
)
block = gr.Blocks(theme=gr.themes.Soft(radius_size=gr.themes.sizes.radius_none, text_size=gr.themes.sizes.text_md))
with block:
with gr.Row():
with gr.Column():
gr.HTML(head)
gr.Markdown(descriptions)
with gr.Accordion(label="🛠️ Instructions:", open=True, elem_id="accordion"):
with gr.Row(equal_height=True):
gr.Markdown(instructions)
first_frame_path_var = gr.State()
tracking_points_var = gr.State([])
with gr.Row():
with gr.Column(scale=1):
image_upload_button = gr.UploadButton(label="Upload Image", file_types=["image"])
add_drag_button = gr.Button(value="Add Drag")
reset_button = gr.Button(value="Reset")
delete_last_drag_button = gr.Button(value="Delete last drag")
delete_last_step_button = gr.Button(value="Delete last step")
with gr.Column(scale=7):
with gr.Row():
with gr.Column(scale=6):
input_image = gr.Image(
label="Input Image",
interactive=True,
height=300,
width=384,
)
with gr.Column(scale=6):
output_image = gr.Image(
label="Motion Path",
interactive=False,
height=256,
width=384,
)
with gr.Row():
with gr.Column(scale=1):
prompt = gr.Textbox(
value="a wonderful elf.",
label="Prompt (highly-recommended)",
interactive=True,
visible=True,
)
negative_prompt = gr.Text(
label="Negative Prompt",
max_lines=5,
placeholder="Please input your negative prompt",
value="worst quality, low quality, letterboxed",
lines=1,
)
drag_mode = gr.Radio(["camera", "object"], label="Drag mode: ", value="object", scale=2)
run_button = gr.Button(value="Run")
with gr.Accordion("More input params", open=False, elem_id="accordion1"):
with gr.Group():
seed = gr.Textbox(label="Seed: ", value=561793204)
randomize_seed = gr.Checkbox(label="Randomize seed", value=False)
with gr.Group():
with gr.Row():
guidance_scale = gr.Slider(
label="Guidance scale",
minimum=1,
maximum=12,
step=0.1,
value=8.5,
)
num_inference_steps = gr.Slider(
label="Number of inference steps",
minimum=1,
maximum=50,
step=1,
value=25,
)
with gr.Group():
personalized = gr.Dropdown(label="Personalized", choices=["", "HelloObject", "TUSUN"], value="")
examples_type = gr.Textbox(label="Examples Type (Ignore) ", value="", visible=False)
with gr.Column(scale=7):
output_video = gr.Video(label="Output Video", width=384, height=256)
with gr.Row():
example = gr.Examples(
label="Input Example",
examples=image_examples,
inputs=[input_image, prompt, drag_mode, seed, personalized, examples_type, output_video],
examples_per_page=10,
cache_examples=False,
)
with gr.Row():
gr.Markdown(citation)
image_upload_button.upload(
preprocess_image,
[image_upload_button, tracking_points_var],
[input_image, first_frame_path_var, tracking_points_var, personalized],
)
add_drag_button.click(add_drag, tracking_points_var, tracking_points_var)
delete_last_drag_button.click(
delete_last_drag,
[tracking_points_var, first_frame_path_var, drag_mode],
[tracking_points_var, input_image],
)
delete_last_step_button.click(
delete_last_step,
[tracking_points_var, first_frame_path_var, drag_mode],
[tracking_points_var, input_image],
)
reset_button.click(
reset_states,
[first_frame_path_var, tracking_points_var],
[input_image, first_frame_path_var, tracking_points_var],
)
input_image.select(
add_tracking_points,
[tracking_points_var, first_frame_path_var, drag_mode],
[tracking_points_var, input_image],
)
run_button.click(
ImageConductor_net.run,
[
first_frame_path_var,
tracking_points_var,
prompt,
drag_mode,
negative_prompt,
seed,
randomize_seed,
guidance_scale,
num_inference_steps,
personalized,
examples_type,
],
[output_image, output_video],
)
block.queue().launch()
