update documentation
Browse files- Ray-Liotta-Goodfellas.jpg +0 -0
- SohoJoeEth + Ray.jpeg +0 -0
- SohoJoeEth.jpeg +0 -0
- app.py +34 -7
Ray-Liotta-Goodfellas.jpg
ADDED
![]() |
SohoJoeEth + Ray.jpeg
ADDED
![]() |
SohoJoeEth.jpeg
ADDED
![]() |
app.py
CHANGED
@@ -228,12 +228,31 @@ with gr.Blocks() as demo:
|
|
228 |
with gr.Row():
|
229 |
gr.Markdown(
|
230 |
"""# Soho-Clip
|
|
|
231 |
|
|
|
|
|
|
|
|
|
232 |
A tool for exploring CLIP embedding spaces.
|
233 |
-
My interest is to use CLIP for image/video understanding (see [CLIP_visual-spatial-reasoning](https://github.com/Sohojoe/CLIP_visual-spatial-reasoning).)
|
234 |
|
235 |
-
Try
|
236 |
""")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
237 |
with gr.Row():
|
238 |
for i in range(max_tabs):
|
239 |
with gr.Tab(f"Input {i+1}"):
|
@@ -287,7 +306,11 @@ Try it out by uploading a few images/add text prompts and generate images of the
|
|
287 |
|
288 |
with gr.Row():
|
289 |
gr.Markdown(
|
290 |
-
"""
|
|
|
|
|
|
|
|
|
291 |
|
292 |
- Combine up to 10 Images and/or text inputs to create an average embedding space.
|
293 |
- View embedding spaces as graph
|
@@ -301,13 +324,17 @@ Try it out by uploading a few images/add text prompts and generate images of the
|
|
301 |
|
302 |
### Acknowledgements
|
303 |
|
304 |
-
- I heavily build on Justin Pinkney's Experiments in Image Variation
|
305 |
-
- [CLIP]
|
306 |
- [Stable Diffusion](https://github.com/CompVis/stable-diffusion)
|
307 |
|
308 |
""")
|
309 |
|
310 |
-
|
|
|
|
|
|
|
|
|
311 |
|
312 |
if __name__ == "__main__":
|
313 |
-
demo.launch()
|
|
|
228 |
with gr.Row():
|
229 |
gr.Markdown(
|
230 |
"""# Soho-Clip
|
231 |
+
""")
|
232 |
|
233 |
+
with gr.Row():
|
234 |
+
with gr.Column(scale=5):
|
235 |
+
gr.Markdown(
|
236 |
+
"""
|
237 |
A tool for exploring CLIP embedding spaces.
|
|
|
238 |
|
239 |
+
Try uploading a few images/add text prompts and click generate images.
|
240 |
""")
|
241 |
+
with gr.Column(scale=3):
|
242 |
+
with gr.Row():
|
243 |
+
with gr.Column(scale=1, min_width=66):
|
244 |
+
gr.Image(value = "SohoJoeEth.jpeg", shape=(66,66), show_label=False, interactive=False).style(height=66, width=66)
|
245 |
+
with gr.Column(scale=1, min_width=15):
|
246 |
+
gr.Markdown("# ")
|
247 |
+
gr.Markdown("# +")
|
248 |
+
with gr.Column(scale=1, min_width=66):
|
249 |
+
gr.Image(value = "Ray-Liotta-Goodfellas.jpg", shape=(66,66), show_label=False, interactive=False).style(height=66, width=66)
|
250 |
+
with gr.Column(scale=1, min_width=15):
|
251 |
+
gr.Markdown("# ")
|
252 |
+
gr.Markdown("# =")
|
253 |
+
with gr.Column(scale=1, min_width=66):
|
254 |
+
gr.Image(value = "SohoJoeEth + Ray.jpeg", shape=(66,66), show_label=False, interactive=False).style(height=66, width=66)
|
255 |
+
|
256 |
with gr.Row():
|
257 |
for i in range(max_tabs):
|
258 |
with gr.Tab(f"Input {i+1}"):
|
|
|
306 |
|
307 |
with gr.Row():
|
308 |
gr.Markdown(
|
309 |
+
"""
|
310 |
+
My interest is to use CLIP for image/video understanding (see [CLIP_visual-spatial-reasoning](https://github.com/Sohojoe/CLIP_visual-spatial-reasoning).)
|
311 |
+
|
312 |
+
|
313 |
+
### Initial Features
|
314 |
|
315 |
- Combine up to 10 Images and/or text inputs to create an average embedding space.
|
316 |
- View embedding spaces as graph
|
|
|
324 |
|
325 |
### Acknowledgements
|
326 |
|
327 |
+
- I heavily build on Justin Pinkney's [Experiments in Image Variation](https://www.justinpinkney.com/image-variation-experiments). Please credit them if you use this work.
|
328 |
+
- [CLIP](https://openai.com/blog/clip/)
|
329 |
- [Stable Diffusion](https://github.com/CompVis/stable-diffusion)
|
330 |
|
331 |
""")
|
332 |
|
333 |
+
# ![Alt Text](file/pup1.jpg)
|
334 |
+
|
335 |
+
# <img src="file/pup1.jpg" width="100" height="100">
|
336 |
+
|
337 |
+
# ![Alt Text](file/pup1.jpg){height=100 width=100}
|
338 |
|
339 |
if __name__ == "__main__":
|
340 |
+
demo.launch()
|