update

app.py

@@ -5,26 +5,26 @@ from anything2image.api import Anything2Image
 anything2img = Anything2Image(imagebind_download_dir='checkpoints')
 
 with gr.Blocks() as demo:
-    gr.HTML(
-            """
-            <div align='center'> <h1>Anything To Image </h1> </div>
-            <p align="center"> Generate image from anything with ImageBind's unified latent space and stable-diffusion-2-1-unclip. </p>
-            <p align="center"><a href="https://github.com/Zeqiang-Lai/Anything2Image"><b>https://github.com/Zeqiang-Lai/Anything2Image</b></p>
-            """)
-    gr.Interface(fn=anything2img, 
-                    inputs=[gr.Text(placeholder="Enter a prompt in addition to the audio, image, text condition below", label="Prompt (Could be empty)"),
-                            "audio", 
-                            "image", 
-                            "text"
-                            ], 
-                    outputs="image",
-                    examples=[['', 'assets/wav/dog_audio.wav', None, None],
-                            ['A painting', 'assets/wav/cat.wav', None, None],
-                            ['', 'assets/wav/wave.wav', 'assets/image/bird.png', None],
-                            ['', None, 'assets/image/bird_image.jpg', None],
-                            ['', None, None, 'A sunset over the ocean.'],
-                            ],
-                    cache_examples=True,
-                    )
+        gr.HTML(
+                """
+                <div align='center'> <h1>Anything To Image </h1> </div>
+                <p align="center"> Generate image from anything with ImageBind's unified latent space and stable-diffusion-2-1-unclip. </p>
+                <p align="center"><a href="https://github.com/Zeqiang-Lai/Anything2Image"><b>https://github.com/Zeqiang-Lai/Anything2Image</b></p>
+                """)
+        gr.Interface(fn=anything2img, 
+                     inputs=[gr.Textbox(placeholder="Enter a prompt in addition to the audio, image, text condition below", label="Prompt (Optional)"),
+                             gr.Audio(label="Audio Condition"), 
+                             gr.Image(label="Image Condition"), 
+                             gr.Textbox(label="Text Condition"),
+                             ], 
+                     outputs=[gr.Image(label="Generated Image")],
+                     examples=[['', 'assets/wav/dog_audio.wav', None, None],
+                               ['A painting', 'assets/wav/cat.wav', None, None],
+                               ['', 'assets/wav/wave.wav', 'assets/image/bird.png', None],
+                               ['', None, 'assets/image/bird_image.jpg', None],
+                               ['', None, None, 'A sunset over the ocean.'],
+                               ],
+                     cache_examples=True,
+                     )
 demo.queue(1).launch()