Spaces:
Running
Running
Update
Browse files
app.py
CHANGED
@@ -1,6 +1,7 @@
|
|
1 |
import gradio as gr
|
2 |
import cv2
|
3 |
import numpy as np
|
|
|
4 |
import os
|
5 |
import torch
|
6 |
import torch.nn.functional as F
|
@@ -12,10 +13,10 @@ from depth_anything.dpt import DepthAnything
|
|
12 |
from depth_anything.util.transform import Resize, NormalizeImage, PrepareForNet
|
13 |
|
14 |
@torch.no_grad()
|
15 |
-
@spaces.GPU
|
16 |
def predict_depth(model, image):
|
17 |
-
return model(image)
|
18 |
|
|
|
19 |
def make_video(video_path, outdir='./vis_video_depth',encoder='vitl'):
|
20 |
if encoder not in ["vitl","vitb","vits"]:
|
21 |
encoder = "vits"
|
@@ -28,7 +29,8 @@ def make_video(video_path, outdir='./vis_video_depth',encoder='vitl'):
|
|
28 |
|
29 |
DEVICE = 'cuda' if torch.cuda.is_available() else 'cpu'
|
30 |
DEVICE = "cuda"
|
31 |
-
depth_anything = DepthAnything.from_pretrained('LiheYoung/depth_anything_{}14'.format(encoder)).to(DEVICE).eval()
|
|
|
32 |
|
33 |
total_params = sum(param.numel() for param in depth_anything.parameters())
|
34 |
print('Total parameters: {:.2f}M'.format(total_params / 1e6))
|
|
|
1 |
import gradio as gr
|
2 |
import cv2
|
3 |
import numpy as np
|
4 |
+
from transformers import pipeline
|
5 |
import os
|
6 |
import torch
|
7 |
import torch.nn.functional as F
|
|
|
13 |
from depth_anything.util.transform import Resize, NormalizeImage, PrepareForNet
|
14 |
|
15 |
@torch.no_grad()
|
|
|
16 |
def predict_depth(model, image):
|
17 |
+
return model(image)["depth"]
|
18 |
|
19 |
+
@spaces.GPU
|
20 |
def make_video(video_path, outdir='./vis_video_depth',encoder='vitl'):
|
21 |
if encoder not in ["vitl","vitb","vits"]:
|
22 |
encoder = "vits"
|
|
|
29 |
|
30 |
DEVICE = 'cuda' if torch.cuda.is_available() else 'cpu'
|
31 |
DEVICE = "cuda"
|
32 |
+
# depth_anything = DepthAnything.from_pretrained('LiheYoung/depth_anything_{}14'.format(encoder)).to(DEVICE).eval()
|
33 |
+
depth_anything = pipeline(task = "depth-estimation", model="nielsr/depth-anything-small", device=0)
|
34 |
|
35 |
total_params = sum(param.numel() for param in depth_anything.parameters())
|
36 |
print('Total parameters: {:.2f}M'.format(total_params / 1e6))
|