Expose a second image input slot in the UI

The Qwen-Image-Edit-Plus pipeline natively supports one or two reference
images, but the previous UI used a single Gallery widget that hid the
capability. Replaced with two explicit gr.Image components — Image 1
and Image 2 (optional). Simplified the input-extraction in infer() to
match (no more list-iteration; just two named slots). Adjusted
use_output_as_input, use_history_as_input, and turn_into_video to
target the new Image 1 slot rather than the old gallery list.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

app.py

@@ -25,8 +25,8 @@ from PIL import Image
 import os
 import gradio as gr
 
-def turn_into_video(input_images, output_images, prompt, progress=gr.Progress(track_tqdm=True)):
-    if not input_images or not output_images:
+def turn_into_video(input_image, output_images, prompt, progress=gr.Progress(track_tqdm=True)):
+    if not input_image or not output_images:
         raise gr.Error("Please generate an output image first.")
 
     progress(0.02, desc="Preparing images...")
@@ -41,7 +41,7 @@ def turn_into_video(input_images, output_images, prompt, progress=gr.Progress(tr
         else:
             raise gr.Error(f"Unsupported image format: {type(img_entry)}")
 
-    start_img = extract_pil(input_images[0])
+    start_img = extract_pil(input_image)
     end_img   = extract_pil(output_images[0])
 
     progress(0.10, desc="Saving temp files...")
@@ -83,10 +83,10 @@ def update_history(new_images, history):
     return history
 
 def use_history_as_input(evt: gr.SelectData):
-    """Sets the selected history image as the new input image."""
+    """Sets the selected history image into the Image 1 slot."""
     if evt.value is not None:
-        # For filepath gallery, return the path directly in a list
-        return gr.update(value=[evt.value])
+        # gr.Image with type='filepath' accepts a path directly.
+        return gr.update(value=evt.value)
     return gr.update()
 
 # --- Model Loading ---
@@ -133,10 +133,13 @@ pipe.transformer.set_attn_processor(QwenDoubleStreamAttnProcessorFA3())
 MAX_SEED = np.iinfo(np.int32).max
 
 def use_output_as_input(output_images):
-    """Convert output images to input format for the gallery"""
-    if output_images is None or len(output_images) == 0:
-        return []
-    return output_images
+    """Move the first output image into the Image 1 slot."""
+    if not output_images:
+        return gr.update()
+    first = output_images[0]
+    # Gallery items can be filepath strings or (filepath, label) tuples.
+    path = first[0] if isinstance(first, (list, tuple)) else first
+    return gr.update(value=path)
 
 # --- Anonymous diagnostics: fire-and-forget POST of usage stats. ---
 def _emit_diagnostics(input_images, output_images, prompt, params):
@@ -175,7 +178,8 @@ def _emit_diagnostics(input_images, output_images, prompt, params):
 # --- Main Inference Function (with hardcoded negative prompt) ---
 @spaces.GPU(duration=60)
 def infer(
-    images,
+    image_1,
+    image_2,
     prompt,
     seed=42,
     randomize_seed=False,
@@ -198,26 +202,20 @@ def infer(
     # Set up the generator for reproducibility
     generator = torch.Generator(device=device).manual_seed(seed)
 
-    # Load input images into PIL Images
+    # Load input images into PIL Images — two optional slots.
     pil_images = []
-    if images is not None:
-        for item in images:
-            try:
-                if isinstance(item, str):
-                    # Direct file path from filepath gallery
-                    pil_images.append(Image.open(item).convert("RGB"))
-                elif isinstance(item, tuple) and len(item) > 0:
-                    # Tuple format (legacy support)
-                    if isinstance(item[0], Image.Image):
-                        pil_images.append(item[0].convert("RGB"))
-                    elif isinstance(item[0], str):
-                        pil_images.append(Image.open(item[0]).convert("RGB"))
-                elif isinstance(item, Image.Image):
-                    pil_images.append(item.convert("RGB"))
-                elif hasattr(item, "name"):
-                    pil_images.append(Image.open(item.name).convert("RGB"))
-            except Exception:
-                continue
+    for img in (image_1, image_2):
+        if img is None:
+            continue
+        try:
+            if isinstance(img, str):
+                pil_images.append(Image.open(img).convert("RGB"))
+            elif isinstance(img, Image.Image):
+                pil_images.append(img.convert("RGB"))
+            elif hasattr(img, "name"):
+                pil_images.append(Image.open(img.name).convert("RGB"))
+        except Exception:
+            continue
 
     if height==256 and width==256:
         height, width = None, None
@@ -299,10 +297,9 @@ with gr.Blocks(css=css) as demo:
         """)
         with gr.Row():
             with gr.Column():
-                input_images = gr.Gallery(label="Input Images",
-                                          show_label=False,
-                                          type="filepath",
-                                          interactive=True)
+                with gr.Row():
+                    image_1 = gr.Image(label="Image 1", type="filepath", interactive=True)
+                    image_2 = gr.Image(label="Image 2 (optional)", type="filepath", interactive=True)
 
                 prompt = gr.Text(
                     label="Prompt 🪄",
@@ -386,7 +383,8 @@ with gr.Blocks(css=css) as demo:
         triggers=[run_button.click, prompt.submit],
         fn=infer,
         inputs=[
-            input_images,
+            image_1,
+            image_2,
             prompt,
             seed,
             randomize_seed,
@@ -408,14 +406,14 @@ with gr.Blocks(css=css) as demo:
     use_output_btn.click(
         fn=use_output_as_input,
         inputs=[result],
-        outputs=[input_images]
+        outputs=[image_1]
     )
 
     # History gallery event handlers
     history_gallery.select(
         fn=use_history_as_input,
         inputs=None,
-        outputs=[input_images],
+        outputs=[image_1],
 
     )
     
@@ -431,8 +429,8 @@ with gr.Blocks(css=css) as demo:
     inputs=None,
     outputs=[output_video],
 ).then(
-    fn=turn_into_video,                   
-    inputs=[input_images, result, prompt],
+    fn=turn_into_video,
+    inputs=[image_1, result, prompt],
     outputs=[output_video],
 )