added camera style presets and controls

camera_presets/angles.json

@@ -0,0 +1,52 @@
+[
+    {
+        "name": "Overhead Shot",
+        "prompt": "{prompt}, (overhead shot:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Top Down",
+        "prompt": "{prompt}, (top down view:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Bird's Eye",
+        "prompt": "{prompt}, (bird's eye view:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "High Angle",
+        "prompt": "{prompt}, (high angle:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Slightly Above",
+        "prompt": "{prompt}, (slightly above:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Straight On",
+        "prompt": "{prompt}, (straight on:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Hero View",
+        "prompt": "{prompt}, (hero view:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Low View",
+        "prompt": "{prompt}, (low view:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Worm Eye View",
+        "prompt": "{prompt}, (worm's eye view:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Selfie",
+        "prompt": "{prompt}, (selfie:{weight})",
+        "default": "1.5"
+    }
+]

camera_presets/distances.json

@@ -0,0 +1,62 @@
+[
+    {
+        "name": "Extreme Close-up",
+        "prompt": "{prompt}, (extreme close-up:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Close-up",
+        "prompt": "{prompt}, (close-up:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Medium Close-up",
+        "prompt": "{prompt}, (medium close-up:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Medium Shot",
+        "prompt": "{prompt}, (medium shot:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Long Shot",
+        "prompt": "{prompt}, (long shot:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Full Shot",
+        "prompt": "{prompt}, (full shot:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Medium Full Shot",
+        "prompt": "{prompt}, (medium full shot:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Establishing Shot",
+        "prompt": "{prompt}, (establishing shot:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Point of View",
+        "prompt": "{prompt}, (point-of-view:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Cowboy Shot",
+        "prompt": "{prompt}, (cowboy shot:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Upper Body",
+        "prompt": "{prompt}, (upper body:{weight})",
+        "default": "1.5"
+    },
+    {
+        "name": "Full Body",
+        "prompt": "{prompt}, (full body:{weight})",
+        "default": "1.5"
+    }
+]

modules/async_worker.py

@@ -1,5 +1,7 @@
 import threading
 
+from modules.camera import apply_camera_preset
+
 
 class AsyncTask:
     def __init__(self, args):
@@ -122,6 +124,10 @@ def handler(async_task):
         prompt = args.pop()
         negative_prompt = args.pop()
         style_selections = args.pop()
+        camera_angle = args.pop()
+        camera_angle_weight = args.pop()
+        camera_distance = args.pop()
+        camera_distance_weight = args.pop()
         performance_selection = args.pop()
         aspect_ratios_selection = args.pop()
         image_number = args.pop()
@@ -162,6 +168,8 @@ def handler(async_task):
             use_expansion = False
 
         use_style = len(style_selections) > 0
+        use_camera_angle = len(camera_angle) > 0
+        use_camera_distance = len(camera_distance) > 0
 
         if base_model_name == refiner_model_name:
             print(f'Refiner disabled because base model and refiner are same.')
@@ -388,6 +396,12 @@ def handler(async_task):
                 positive_basic_workloads = []
                 negative_basic_workloads = []
 
+                if use_camera_angle:
+                    task_prompt = apply_camera_preset(preset='angle', name=camera_angle, weight=camera_angle_weight, positive=task_prompt)
+
+                if use_camera_distance:
+                    task_prompt = apply_camera_preset(preset='distance', name=camera_distance, weight=camera_distance_weight, positive=task_prompt)
+
                 if use_style:
                     for s in style_selections:
                         p, n = apply_style(s, positive=task_prompt)
@@ -775,11 +789,13 @@ def callback(step, x0, x, total_steps, y):
                     imgs = [inpaint_worker.current_task.post_process(x) for x in imgs]
 
                 for x in imgs:
+                    camera_settings = str([x for x in [camera_angle, camera_distance] if len(x) > 0])
+                    camera_settings = len(camera_settings) > 0 and " | " + camera_settings or ''
                     d = [
                         ('Prompt', task['log_positive_prompt']),
                         ('Negative Prompt', task['log_negative_prompt']),
                         ('Fooocus V2 Expansion', task['expansion']),
-                        ('Styles', str(raw_style_selections)),
+                        ('Styles', str(raw_style_selections) + camera_settings),
                         ('Performance', performance_selection),
                         ('Resolution', str((width, height))),
                         ('Sharpness', sharpness),

modules/camera.py

@@ -0,0 +1,98 @@
+import json
+import os
+
+from modules.util import normalize_key
+
+
+def load_presets(file_path):
+    """
+    This function loads presets from a given file path. It reads the file, parses the JSON content, and stores the presets in a dictionary.
+    Each preset is a tuple containing a prompt and a weight.
+
+    Args:
+        file_path (str): The path to the file containing the presets.
+
+    Returns:
+        dict: A dictionary where the keys are the normalized names of the presets and the values are tuples containing the prompt and the weight of each preset.
+    """
+    presets = {}
+    try:
+        with open(file_path, encoding='utf-8') as f:
+            for preset in json.load(f):
+                key = normalize_key(preset['name'])
+                prompt = preset.get('prompt', '')
+                weight = preset.get('default', 1)
+                presets[key] = (prompt, weight)
+    except Exception as e:
+        print(str(e))
+        print(f'Failed to load camera presets file {file_path}')
+    return presets
+
+
+# Paths to the files containing the camera angles and distances presets
+angles_path = os.path.join(os.path.dirname(__file__), '../camera_presets/angles.json')
+distances_path = os.path.join(os.path.dirname(__file__), '../camera_presets/distances.json')
+
+# Load the camera angles and distances presets
+camera_angles = load_presets(angles_path)
+camera_distances = load_presets(distances_path)
+
+# Get the names of the camera angles and distances presets for use in the UI
+camera_angle_names = list(camera_angles.keys())
+camera_distance_names = list(camera_distances.keys())
+
+
+def get_preset_weight(preset, name):
+    """
+    This function retrieves the weight of a given preset.
+
+    Args:
+        preset (str): The type of the preset ('angle' or 'distance').
+        name (str): The name of the preset.
+
+    Returns:
+        str: The weight of the preset.
+    """
+    _, w = _get_preset(preset, name)
+    return w
+
+
+def apply_camera_preset(preset, name, weight, positive):
+    """
+    This function applies a camera preset. It replaces the placeholders in the presets' prompt with the given weight and positive value.
+
+    Args:
+        preset (str): The type of the preset ('angle' or 'distance').
+        name (str): The name of the preset.
+        weight (str): The weight to replace the '{weight}' placeholder in the presets' prompt.
+        positive (str): The value to replace the '{prompt}' placeholder in the presets' prompt.
+
+    Returns:
+        str: The presets' prompt with the placeholders replaced by the given weight and positive value.
+    """
+    p, _ = _get_preset(preset, name)
+    return p.replace('{weight}', str(weight)).replace('{prompt}', positive)
+
+
+def _get_preset(preset, name):
+    """
+    This function retrieves a preset from the appropriate dictionary based on the preset type.
+
+    Args:
+        preset (str): The type of the preset ('angle' or 'distance').
+        name (str): The name of the preset.
+
+    Returns:
+        tuple: The preset (a tuple containing the prompt and the weight).
+
+    Raises:
+        ValueError: If the preset type is not 'angle' or 'distance'.
+    """
+    if preset == 'angle':
+        camera_presets = camera_angles
+    elif preset == 'distance':
+        camera_presets = camera_distances
+    else:
+        raise ValueError(f'Unknown preset type : {preset}')
+
+    return camera_presets[name]

modules/sdxl_styles.py

@@ -2,7 +2,7 @@
 import re
 import json
 
-from modules.util import get_files_from_folder
+from modules.util import get_files_from_folder, normalize_key
 
 
 # cannot use modules.config - validators causing circular imports
@@ -11,11 +11,8 @@
 wildcards_max_bfs_depth = 64
 
 
-def normalize_key(k):
-    k = k.replace('-', ' ')
-    words = k.split(' ')
-    words = [w[:1].upper() + w[1:].lower() for w in words]
-    k = ' '.join(words)
+def normalize_style_key(k):
+    k = normalize_key(k)
     k = k.replace('3d', '3D')
     k = k.replace('Sai', 'SAI')
     k = k.replace('Mre', 'MRE')
@@ -41,7 +38,7 @@ def normalize_key(k):
     try:
         with open(os.path.join(styles_path, styles_file), encoding='utf-8') as f:
             for entry in json.load(f):
-                name = normalize_key(entry['name'])
+                name = normalize_style_key(entry['name'])
                 prompt = entry['prompt'] if 'prompt' in entry else ''
                 negative_prompt = entry['negative_prompt'] if 'negative_prompt' in entry else ''
                 styles[name] = (prompt, negative_prompt)

modules/util.py

@@ -175,3 +175,21 @@ def get_files_from_folder(folder_path, exensions=None, name_filter=None):
                 filenames.append(path)
 
     return sorted(filenames, key=lambda x: -1 if os.sep in x else 1)
+
+
+def normalize_key(k):
+    """
+    This function takes a string key as input and transforms it to be more human.
+
+    Args:
+        k (str): The string to be normalized.
+
+    Returns:
+        str: The normalized string.
+    """
+    # Replace hyphens with spaces
+    # Split the string into words
+    # Capitalize the first letter of each word and make the rest of the letters lowercase
+    # Join the words back together with spaces in between
+
+    return ' '.join(w.capitalize() for w in k.replace('-', ' ').split())

webui.py

@@ -18,6 +18,7 @@
 import copy
 
 from modules.sdxl_styles import legal_style_names
+from modules.camera import camera_angle_names, camera_distance_names, get_preset_weight
 from modules.private_logger import get_current_html_path
 from modules.ui_gradio_extensions import reload_javascript
 from modules.auth import auth_enabled, check_auth
@@ -230,6 +231,26 @@ def ip_advance_checked(x):
                                                    value=modules.config.default_aspect_ratio, info='width × height',
                                                    elem_classes='aspect_ratios')
                 image_number = gr.Slider(label='Image Number', minimum=1, maximum=modules.config.default_max_image_number, step=1, value=modules.config.default_image_number)
+
+                with gr.Group():
+                    # TODO: refactor UI
+                    with gr.Row():
+                        camera_angle = gr.Dropdown(label=f'Camera Angle',
+                                                 choices=camera_angle_names, value=None, scale=2)
+                        camera_angle_weight = gr.Number(label='Weight', minimum=0, maximum=2, step=0.01, value=0, scale=1)
+
+                        camera_angle.change(lambda x: camera_angle_weight.update(value=get_preset_weight('angle', x)),
+                                            inputs=[camera_angle], outputs=[camera_angle_weight],
+                                            queue=False, show_progress=False)
+
+                    with gr.Row():
+                        camera_distance = gr.Dropdown(label=f'Camera Distance', choices=camera_distance_names, value=None, scale=2)
+                        camera_distance_weight = gr.Number(label='Weight', minimum=0, maximum=2, step=0.01, value=0, scale=1)
+
+                        camera_distance.change(lambda x: camera_distance_weight.update(value=get_preset_weight('distance', x)),
+                                            inputs=[camera_distance], outputs=[camera_distance_weight],
+                                            queue=False, show_progress=False)
+
                 negative_prompt = gr.Textbox(label='Negative Prompt', show_label=True, placeholder="Type prompt here.",
                                              info='Describing what you do not want to see.', lines=2,
                                              elem_id='negative_prompt',
@@ -321,6 +342,7 @@ def refresh_seed(r, seed_string):
 
                 with gr.Row():
                     model_refresh = gr.Button(label='Refresh', value='\U0001f504 Refresh All Files', variant='secondary', elem_classes='refresh_button')
+
             with gr.Tab(label='Advanced'):
                 guidance_scale = gr.Slider(label='Guidance Scale', minimum=1.0, maximum=30.0, step=0.01,
                                            value=modules.config.default_cfg_scale,
@@ -521,6 +543,7 @@ def inpaint_mode_change(mode):
 
         ctrls = [
             prompt, negative_prompt, style_selections,
+            camera_angle, camera_angle_weight, camera_distance, camera_distance_weight,
             performance_selection, aspect_ratios_selection, image_number, image_seed, sharpness, guidance_scale
         ]
 
@@ -559,6 +582,10 @@ def parse_meta(raw_prompt_txt, is_generating):
             prompt,
             negative_prompt,
             style_selections,
+            camera_angle,
+            camera_angle_weight,
+            camera_distance,
+            camera_distance_weight,
             performance_selection,
             aspect_ratios_selection,
             overwrite_width,