Update app.py
Browse files
app.py
CHANGED
@@ -82,8 +82,14 @@ def save_splat_file(splat_data, output_path):
|
|
82 |
with open(output_path, "wb") as f:
|
83 |
f.write(splat_data)
|
84 |
|
85 |
-
def get_reconstructed_scene(outdir,
|
86 |
|
|
|
|
|
|
|
|
|
|
|
|
|
87 |
images = [process_image(img_path) for img_path in image_files]
|
88 |
images = torch.stack(images, dim=0).unsqueeze(0).to(device) # [1, K, 3, 448, 448]
|
89 |
b, v, c, h, w = images.shape
|
@@ -234,12 +240,12 @@ def generate_splats_from_video(video_path, session_id=None):
|
|
234 |
session_id = uuid.uuid4().hex
|
235 |
|
236 |
images_folder, image_paths = extract_frames(video_path, session_id)
|
237 |
-
plyfile, rgb_vid, depth_vid = generate_splats_from_images(
|
238 |
|
239 |
return plyfile, rgb_vid, depth_vid, image_paths
|
240 |
|
241 |
@spaces.GPU()
|
242 |
-
def generate_splats_from_images(
|
243 |
|
244 |
if session_id is None:
|
245 |
session_id = uuid.uuid4().hex
|
@@ -250,9 +256,16 @@ def generate_splats_from_images(image_paths, session_id=None):
|
|
250 |
|
251 |
base_dir = os.path.join(os.environ["ANYSPLAT_PROCESSED"], session_id)
|
252 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
253 |
print("Running run_model...")
|
254 |
with torch.no_grad():
|
255 |
-
plyfile, video, depth_colored = get_reconstructed_scene(base_dir,
|
256 |
|
257 |
end_time = time.time()
|
258 |
print(f"Total time: {end_time - start_time:.2f} seconds (including IO)")
|
@@ -322,7 +335,7 @@ if __name__ == "__main__":
|
|
322 |
with gr.Tab("Video"):
|
323 |
input_video = gr.Video(label="Upload Video", sources=["upload"], interactive=True, height=512)
|
324 |
with gr.Tab("Images"):
|
325 |
-
input_images = gr.File(file_count="multiple", label="Upload Files",
|
326 |
|
327 |
submit_btn = gr.Button(
|
328 |
"Generate Gaussian Splat", scale=1, variant="primary"
|
@@ -400,7 +413,7 @@ if __name__ == "__main__":
|
|
400 |
|
401 |
submit_btn.click(
|
402 |
fn=generate_splats_from_images,
|
403 |
-
inputs=[
|
404 |
outputs=[reconstruction_output, rgb_video, depth_video])
|
405 |
|
406 |
input_video.upload(
|
@@ -417,4 +430,4 @@ if __name__ == "__main__":
|
|
417 |
|
418 |
demo.unload(cleanup)
|
419 |
demo.queue()
|
420 |
-
demo.launch(show_error=True, share=True)
|
|
|
82 |
with open(output_path, "wb") as f:
|
83 |
f.write(splat_data)
|
84 |
|
85 |
+
def get_reconstructed_scene(outdir, model, device):
|
86 |
|
87 |
+
image_files = sorted(
|
88 |
+
[
|
89 |
+
os.path.join(outdir, "images", f)
|
90 |
+
for f in os.listdir(os.path.join(outdir, "images"))
|
91 |
+
]
|
92 |
+
)
|
93 |
images = [process_image(img_path) for img_path in image_files]
|
94 |
images = torch.stack(images, dim=0).unsqueeze(0).to(device) # [1, K, 3, 448, 448]
|
95 |
b, v, c, h, w = images.shape
|
|
|
240 |
session_id = uuid.uuid4().hex
|
241 |
|
242 |
images_folder, image_paths = extract_frames(video_path, session_id)
|
243 |
+
plyfile, rgb_vid, depth_vid = generate_splats_from_images(images_folder, session_id)
|
244 |
|
245 |
return plyfile, rgb_vid, depth_vid, image_paths
|
246 |
|
247 |
@spaces.GPU()
|
248 |
+
def generate_splats_from_images(images_folder, session_id=None):
|
249 |
|
250 |
if session_id is None:
|
251 |
session_id = uuid.uuid4().hex
|
|
|
256 |
|
257 |
base_dir = os.path.join(os.environ["ANYSPLAT_PROCESSED"], session_id)
|
258 |
|
259 |
+
all_files = (
|
260 |
+
sorted(os.listdir(images_folder))
|
261 |
+
if os.path.isdir(images_folder)
|
262 |
+
else []
|
263 |
+
)
|
264 |
+
all_files = [f"{i}: {filename}" for i, filename in enumerate(all_files)]
|
265 |
+
|
266 |
print("Running run_model...")
|
267 |
with torch.no_grad():
|
268 |
+
plyfile, video, depth_colored = get_reconstructed_scene(base_dir, model, device)
|
269 |
|
270 |
end_time = time.time()
|
271 |
print(f"Total time: {end_time - start_time:.2f} seconds (including IO)")
|
|
|
335 |
with gr.Tab("Video"):
|
336 |
input_video = gr.Video(label="Upload Video", sources=["upload"], interactive=True, height=512)
|
337 |
with gr.Tab("Images"):
|
338 |
+
input_images = gr.File(file_count="multiple", label="Upload Files", height=512)
|
339 |
|
340 |
submit_btn = gr.Button(
|
341 |
"Generate Gaussian Splat", scale=1, variant="primary"
|
|
|
413 |
|
414 |
submit_btn.click(
|
415 |
fn=generate_splats_from_images,
|
416 |
+
inputs=[target_dir_output, session_state],
|
417 |
outputs=[reconstruction_output, rgb_video, depth_video])
|
418 |
|
419 |
input_video.upload(
|
|
|
430 |
|
431 |
demo.unload(cleanup)
|
432 |
demo.queue()
|
433 |
+
demo.launch(show_error=True, share=True)
|