Initial push of StyleSavvy AI stylist app

.gitignore

@@ -0,0 +1,5 @@
+__pycache__/
+.env
+venv/
+*.pyc
+.DS_Store

app.py

@@ -0,0 +1,76 @@
+import gradio as gr
+from utils.detector import detect_clothing
+from utils.advisor import get_advice
+
+def run_style_savvy(image, bg_remove, body_type, face_shape, gender, occasion):
+    items = detect_clothing(image, do_bg_remove=bg_remove)
+    advice_list = get_advice(items, body_type, face_shape, gender, occasion)
+
+    unique_advice = []
+    seen = set()
+    for tip in advice_list:
+        if tip not in seen:
+            unique_advice.append(tip)
+            seen.add(tip)
+
+    html = """
+    <div style="
+        background-color: #1e1e1e;
+        color: #f5f5f5;
+        padding: 24px;
+        border-radius: 16px;
+        max-width: 640px;
+        margin: auto;
+        font-family: 'Segoe UI', sans-serif;
+    ">
+      <h2 style="
+          margin-top: 0;
+          font-size: 2em;
+          color: #ff8c00;
+          text-align: center;
+          text-transform: uppercase;
+      ">
+        ✨ Your Personalized Style Tips ✨
+      </h2>
+      <ol style="
+          padding-left: 20px;
+          font-size: 1.2em;
+          line-height: 1.8;
+      ">
+    """
+    for advice in unique_advice:
+        html += f"<li style='margin-bottom: 12px;'>{advice}</li>"
+    html += "</ol></div>"
+
+    return html
+
+with gr.Blocks(theme=gr.themes.Soft()) as demo:
+    gr.Markdown("## 👗 StyleSavvy — AI Fashion Consultant")
+    gr.Markdown("Upload your photo and get personalized fashion advice tailored to your features and occasion.")
+
+    with gr.Row():
+        with gr.Column(scale=1):
+            with gr.Group():
+                gr.Markdown("### 🧾 Style Details")
+                bg_remove = gr.Checkbox(label="Remove Background")
+                body_type = gr.Radio(["Slim", "Athletic", "Curvy", "Plus-size"], label="Body Type")
+                face_shape = gr.Radio(["Oval", "Round", "Square", "Heart"], label="Face Shape")
+                gender = gr.Radio(["Male", "Female"], label="Gender")
+                occasion = gr.Textbox(label="Occasion", placeholder="e.g. Wedding, Office Party")
+
+        with gr.Column(scale=1):
+            with gr.Group():
+                gr.Markdown("### 📸 Upload Your Look")
+                image = gr.Image(type="pil", label="Upload Photo")
+
+    submit_btn = gr.Button("✨ Generate Style Tips")
+    output = gr.HTML()
+
+    submit_btn.click(
+        fn=run_style_savvy,
+        inputs=[image, bg_remove, body_type, face_shape, gender, occasion],
+        outputs=output
+    )
+
+if __name__ == "__main__":
+    demo.launch()

models/llm.py

@@ -0,0 +1,83 @@
+
+from transformers import pipeline
+from typing import List
+
+PROMPTS = {
+    "category_expansion": (
+        "As a top-tier fashion advisor, craft one impactful styling suggestion for a {gender} individual with a {body_type} body "
+        "and {face_shape} face attending the {occasion}. They have on {items}. "
+        "Highlight a strategic enhancement in silhouette, color scheme, accessory choice, or footwear to elevate their look."
+    ),
+    "event_aesthetic": (
+        "Imagine you are curating an immersive style experience for a {gender} attendee with a {body_type} silhouette and {face_shape} face at the {occasion}. "
+        "They’re currently wearing {items}. Provide one highly descriptive recommendation that harmonizes fabric textures, color temperature, silhouette, and accessory accents with the event’s specific ambiance, lighting conditions, and seasonal atmosphere."
+    ),
+    "fashion_editor": (
+        "You are the Editor-in-Chief of a prestigious fashion publication. Advise a {gender} trendsetter with a {body_type} frame and {face_shape} face attending the {occasion}, "
+        "currently in {items}. Offer one magazine-cover-worthy styling tip—highlight a trending color palette, editorial-worthy silhouette, and innovative accessory placement that will resonate with a discerning audience."
+    ),
+    "influencer_style": (
+        "As a cutting-edge style influencer with millions of followers, recommend one eye-catching flair tip for a {gender} follower with a {body_type} physique and {face_shape} face, "
+        "heading to the {occasion} in {items}. Frame it as a social-media-caption-ready moment: mention a statement accessory, bold color pop, or texture twist that will go viral."
+    ),
+    "seasonal_trend": (
+        "As a seasonal style expert specializing in spring/summer trends, guide a {gender} individual with a {body_type} shape and {face_shape} face preparing for the {occasion}. "
+        "They currently wear {items}. Provide one tip incorporating current seasonal motifs—think floral prints, breathable linens, or eco-friendly fabrics—that elevates their ensemble."
+    ),
+}
+
+class StyleSavvy:
+    def __init__(
+        self,
+        model_name: str = "google/flan-t5-large",
+        device: int   = -1,     # -1 = CPU, or GPU index
+        max_length: int = 150,
+    ):
+        # A local instruction-tuned T5 model
+        self.pipe = pipeline(
+            "text2text-generation",
+            model=model_name,
+            tokenizer=model_name,
+            device=device,
+        )
+        self.max_length = max_length
+        self.num_beams = 4
+    # TODO: Modification: Add more prompts to the advise function
+    # to make it more specific to the user's needs.
+    # The function now takes in the user's body type, face shape, and occasion
+    # and generates style tips accordingly.
+
+    def advise(self,
+               items: List[str],
+               body_type: str,
+               face_shape: str,
+               gender: str,
+               occasion: str
+    ) -> List[str]:
+        """
+        Generate one result per prompt template and return all as a list.
+        """
+        labels = ", ".join(items) if items else "an outfit"
+        results: List[str] = []
+        for tpl in PROMPTS.values():
+            prompt = tpl.format(
+                body_type=body_type,
+                face_shape=face_shape,
+                gender = gender,
+                occasion=occasion,
+                items=labels
+            )
+            out = self.pipe(
+                prompt,
+                max_length=self.max_length,
+                num_beams=self.num_beams,
+                early_stopping=True,
+                do_sample=False,
+                no_repeat_ngram_size=3,  # avoid repeating phrases
+            )[0]["generated_text"].strip()
+            results.append(out)
+        return results
+
+
+
+

models/vision.py

@@ -0,0 +1,50 @@
+
+# models/vision.py -- Working
+
+from transformers import pipeline
+from PIL import Image
+
+class VisionModel:
+    def __init__(
+        self,
+        model_name: str = "valentinafeve/yolos-fashionpedia",
+        threshold: float = 0.7
+    ):
+        self.pipe = pipeline("object-detection", model=model_name)
+        self.threshold = threshold
+
+    def detect(self, image: Image.Image):
+        # 1) Ensure RGB
+        if image.mode != "RGB":
+            image = image.convert("RGB")
+
+        # 2) Run detection
+        results = self.pipe(image)
+
+        # 3) Process & filter
+        processed = []
+        for r in results:
+            score = float(r["score"])
+            if score < self.threshold:
+                continue
+
+            # r["box"] is a dict: {"xmin":..., "ymin":..., "xmax":..., "ymax":...}
+            box = r["box"]
+            coords = [
+                float(box["xmin"]),
+                float(box["ymin"]),
+                float(box["xmax"]),
+                float(box["ymax"]),
+            ]
+
+            processed.append({
+                "label": r["label"],
+                "score": score,
+                "box": coords
+            })
+
+        return processed
+
+
+
+

requirements.txt

@@ -0,0 +1,71 @@
+aiofiles==24.1.0
+annotated-types==0.7.0
+anyio==4.9.0
+certifi==2025.4.26
+charset-normalizer==3.4.1
+click==8.1.8
+fastapi==0.115.12
+ffmpy==0.5.0
+filelock==3.18.0
+fsspec==2025.3.2
+gradio==5.28.0
+gradio_client==1.10.0
+groovy==0.1.2
+h11==0.16.0
+httpcore==1.0.9
+httpx==0.28.1
+huggingface-hub==0.30.2
+idna==3.10
+inquirerpy==0.3.4
+Jinja2==3.1.6
+markdown-it-py==3.0.0
+MarkupSafe==3.0.2
+mdurl==0.1.2
+mpmath==1.3.0
+networkx==3.4.2
+numpy==2.2.5
+orjson==3.10.18
+packaging==25.0
+pandas==2.2.3
+pfzy==0.3.4
+pillow==11.2.1
+prompt_toolkit==3.0.51
+protobuf==6.30.2
+pydantic==2.11.4
+pydantic_core==2.33.2
+pydub==0.25.1
+Pygments==2.19.1
+python-dateutil==2.9.0.post0
+python-dotenv==1.1.0
+python-multipart==0.0.20
+pytz==2025.2
+PyYAML==6.0.2
+regex==2024.11.6
+requests==2.32.3
+rich==14.0.0
+ruff==0.11.7
+safehttpx==0.1.6
+safetensors==0.5.3
+semantic-version==2.10.0
+sentencepiece==0.2.0
+setuptools==80.0.1
+shellingham==1.5.4
+six==1.17.0
+sniffio==1.3.1
+starlette==0.46.2
+sympy==1.14.0
+timm==1.0.15
+tokenizers==0.21.1
+tomlkit==0.13.2
+torch==2.7.0
+torchvision==0.22.0
+tqdm==4.67.1
+transformers==4.51.3
+typer==0.15.3
+typing-inspection==0.4.0
+typing_extensions==4.13.2
+tzdata==2025.2
+urllib3==2.4.0
+uvicorn==0.34.2
+wcwidth==0.2.13
+websockets==15.0.1

utils/advisor.py

@@ -0,0 +1,7 @@
+from models.llm import StyleSavvy
+
+advisor = StyleSavvy()
+
+def get_advice(items, body_type, face_shape, gender,occasion):
+    return advisor.advise(items, body_type, face_shape, gender, occasion)
+

utils/bg_removal.py

@@ -0,0 +1,20 @@
+import os, requests
+from io import BytesIO
+from PIL import Image
+from dotenv import load_dotenv
+
+load_dotenv()
+
+API_KEY = os.getenv("REMOVE_BG_API_KEY")
+ENDPOINT = "https://api.remove.bg/v1.0/removebg"
+
+def remove_background(image_bytes: bytes) -> Image.Image:
+    resp = requests.post(
+        ENDPOINT,
+        files ={"image_file": ("image.jpg", image_bytes, "image/jpeg")},
+        data = {"size": "auto"},
+        headers = {"X-Api-Key": API_KEY},
+    )
+    resp.raise_for_status()
+    return Image.open(BytesIO(resp.content))
+

utils/detector.py

@@ -0,0 +1,59 @@
+
+from io import BytesIO
+from PIL import Image
+from models.vision import VisionModel
+from utils.bg_removal import remove_background
+
+vision = VisionModel()
+FASHION_LABELS = {
+    "shirt", "t-shirt", "blouse", "tank top", "sweater", "hoodie", "jacket",
+    "coat", "overcoat", "raincoat", "windbreaker", "cardigan", "blazer",
+    "pants", "jeans", "shorts", "leggings", "tights", "skirt", "dress",
+    "suit", "jumpsuit", "romper", "vest", "sports bra", "tracksuit",
+    "belt", "tie", "scarf", "hat", "cap", "gloves", "socks",
+    "shoe", "sneakers", "boots", "sandals", "heels",
+    "watch", "necklace", "bracelet", "earrings", "ring",
+    "backpack", "handbag", "purse", "wallet"
+}
+
+def detect_clothing(image_input, do_bg_remove: bool = False):
+    # 1) Load into a PIL.Image if it's a filepath
+    if isinstance(image_input, str):
+        img = Image.open(image_input)
+    else:
+        img = image_input
+
+    # 2) Optionally remove background (works on bytes)
+    if do_bg_remove:
+        buf = BytesIO()
+        img.convert("RGB").save(buf, format="JPEG")
+        img_bytes = buf.getvalue()
+        img = remove_background(img_bytes)
+    else:
+        # ensure you drop any alpha channel
+        img = img.convert("RGB")
+
+    # 3) Run detection
+    raw_detections = vision.detect(img)
+
+    # 4) Filter and deduplicate
+    filtered = {}
+    for det in raw_detections:
+        label = det["label"].lower()
+        if label in FASHION_LABELS:
+            # Only keep the first or highest score if multiple detected
+            if label not in filtered or det["score"] > filtered[label]["score"]:
+                filtered[label] = {
+                    "label": label,
+                    "score": det["score"],
+                    "box": det.get("box", [])
+                }
+
+    # 5) Return dict or fallback if empty
+    if not filtered:
+        return {"outfit": {"label": "outfit", "score": 1.0, "box": []}}
+
+    return filtered
+
+
+

utils/test_detector.py

@@ -0,0 +1,49 @@
+# test_detector.py
+
+from detector import detect_clothing
+from PIL import Image, ImageDraw
+import os
+
+def visualize_and_print(image_path, do_bg_remove=False, output_dir="vis"):
+    # Ensure output folder exists
+    os.makedirs(output_dir, exist_ok=True)
+
+    img = Image.open(image_path).convert("RGB")
+    print(f"\n--- Testing {os.path.basename(image_path)} (bg_remove={do_bg_remove}) ---")
+    
+    # Run your detector
+    dets = detect_clothing(img, do_bg_remove=do_bg_remove)
+    if not dets:
+        print("No detections!")
+        return
+
+    # Print raw detections
+    # Print raw detections
+    for i, d in enumerate(dets.values(), 1):
+        lbl = d["label"]
+        scr = d["score"]
+        box = d.get("box", [])
+        print(f" {i}. {lbl:12s} @ {scr:.2f} → {box}")
+
+    # Draw boxes
+    vis = img.copy()
+    draw = ImageDraw.Draw(vis)
+    for d in dets.values():
+        if d.get("box"):
+            x0, y0, x1, y1 = d["box"]
+            draw.rectangle([x0, y0, x1, y1], outline="red", width=2)
+            draw.text((x0, y0 - 10), f"{d['label']}:{d['score']:.2f}", fill="red")
+        # Save visualization
+        out_path = os.path.join(output_dir, os.path.basename(image_path))
+        vis.save(out_path)
+        print(f" Visualization saved to {out_path}")
+    
+if __name__ == "__main__":
+    # List your test images here
+    samples = [
+        "/Users/tanzimfarhan/Desktop/Python/Codes/SLU/CS5930/FinalProject/StyleSavvy/images/casual.jpg",
+        "/Users/tanzimfarhan/Desktop/Python/Codes/SLU/CS5930/FinalProject/StyleSavvy/images/WomenCasual.jpg",
+    ]
+    for img_path in samples:
+        visualize_and_print(img_path, do_bg_remove=False)
+        # visualize_and_print(img_path, do_bg_remove=True)

utils/test_llm.py

@@ -0,0 +1,193 @@
+# # test_llm.py
+# """
+# Test harness for StyleSavvy LLM prompts.
+# Defines multiple prompt templates and evaluates the generated outputs,
+# checking for the expected number of bullet-point style tips.
+# """
+# from models.llm import StyleSavvy
+
+# # Variant prompt templates with placeholders
+# PROMPT_TEMPLATES = {
+#     "occasion_driven": (
+#         "You are an expert fashion stylist. A client is preparing for {occasion}. "
+#         "They have a {body_type}-shaped body and a {face_shape} face. They’re currently wearing: {items}. "
+#         "Give 3 to 5 *distinct* style tips focused on making them look their best at the event. "
+#         "Make the suggestions relevant to the setting, weather, and formality of the occasion. "
+#         "Avoid repeating any advice."
+#     ),
+
+#     "function_based": (
+#         "You're advising someone with a {body_type} build and {face_shape} face. "
+#         "They're attending a {occasion} and are wearing {items}. "
+#         "Suggest 3–5 concise fashion improvements or enhancements. "
+#         "Each suggestion should be unique and tailored to the event. "
+#         "Include practical choices for color, layering, accessories, or footwear. "
+#         "Avoid repeating words or phrases."
+#     ),
+
+#     "intent_style": (
+#         "Act as a high-end personal stylist. Your client has a {body_type} body shape and a {face_shape} face. "
+#         "They're going to a {occasion} and are wearing {items}. "
+#         "Write 3 to 5 brief but powerful styling suggestions to elevate their look. "
+#         "Focus on intent—what feeling or impression each style choice creates for the event."
+#     ),
+# }
+
+
+# # Test parameters
+# BODY_TYPE = "Slim"
+# FACE_SHAPE = "Round"
+# OCCASION = "Rooftop Evening Party"
+# ITEMS = ["shirt", "jeans", "jacket","shoes"]
+
+# if __name__ == "__main__":
+#     advisor = StyleSavvy()
+
+#     for name, template in PROMPT_TEMPLATES.items():
+#         # Build prompt by replacing placeholders
+#         prompt = template.format(
+#             body_type=BODY_TYPE,
+#             face_shape=FACE_SHAPE,
+#             occasion=OCCASION,
+#             items=", ".join(ITEMS)
+#         )
+#         print(f"=== Testing template: {name} ===")
+#         print("Prompt:")
+#         print(prompt)
+
+#         # Generate output (use only supported args)
+#         result = advisor.pipe(
+#             prompt,
+#             max_length=advisor.max_length,
+#             early_stopping=True,
+#             do_sample=False
+#         )[0]["generated_text"].strip()
+
+#         print("Generated output:")
+#         print(result)
+
+#         # Extract bullet lines
+#         bullets = [ln for ln in result.splitlines() if ln.strip().startswith("- ")]
+#         print(f"Number of bullets detected: {len(bullets)}")
+#         for i, b in enumerate(bullets, start=1):
+#             print(f" {i}. {b}")
+#         print("" + "-"*40)
+
+
+# test_llm.py
+"""
+Test harness for StyleSavvy LLM prompts.
+Evaluates multiple prompt templates and parses the generated outputs into distinct tips.
+"""
+
+from models.llm import StyleSavvy
+
+# Variant prompt templates with placeholders
+# PROMPTS = {
+#     "direct_instruction": (
+#         "You are a professional fashion stylist. A client with a {body_type}-shaped body "
+#         "and {face_shape} face is preparing for {occasion}. They are currently wearing: {items}. "
+#         "Give exactly five distinct styling tips to improve their outfit. "
+#         "Each tip should be concise, actionable, and start on a new line."
+#     ),
+#     "category_expansion": (
+#         "As a high-end fashion advisor, provide five styling tips for a {body_type}-shaped person "
+#         "with a {face_shape} face attending {occasion}. They are wearing {items}. "
+#         "Offer one tip each for silhouette, color, accessories, footwear, and layering, "
+#         "each on its own line."
+#     ),
+#     "event_aesthetic": (
+#         "Imagine curating the perfect outfit for a {body_type}-shaped individual with a {face_shape} face "
+#         "at {occasion}. They are wearing {items}. Suggest 5 ways to enhance their style, "
+#         "focusing on event-appropriate aesthetics. Separate each tip with a newline."
+#     ),
+#     "fashion_editor": (
+#         "As a fashion editor, outline five unique styling tips for a {body_type}-shaped reader with a {face_shape} face "
+#         "attending {occasion}. They wear {items}. Each recommendation should reflect expertise and relevance. "
+#         "List each tip on a new line."
+#     ),
+#     "influencer_style": (
+#         "You’re an influencer giving sharp styling advice. A follower with a {body_type} body and {face_shape} face "
+#         "is going to {occasion}, wearing {items}. Reply with five snappy, modern style tips, "
+#         "each on its own line."
+#     ),
+# }
+
+PROMPTS = {
+    "direct_instruction": (
+        "You are a world-renowned fashion stylist celebrated for your bold creativity and attention to detail. "
+        "Your {gender} client has a {body_type}-shaped silhouette and a {face_shape} face, preparing for the {occasion}. "
+        "They’re wearing {items}. In vivid, sensory-rich language, provide one transformative styling recommendation that considers the event’s ambiance, lighting, and dress code. "
+        "Use dynamic adjectives and actionable insight to elevate their entire look."
+    ),
+    "category_expansion": (
+        "As a top-tier fashion advisor, craft one impactful styling suggestion for a {gender} individual with a {body_type} body "
+        "and {face_shape} face attending the {occasion}. They have on {items}. "
+        "Highlight a strategic enhancement in silhouette, color scheme, accessory choice, or footwear to elevate their look."
+    ),
+    "event_aesthetic": (
+        "Imagine you are curating an immersive style experience for a {gender} attendee with a {body_type} silhouette and {face_shape} face at the {occasion}. "
+        "They’re currently wearing {items}. Provide one highly descriptive recommendation that harmonizes fabric textures, color temperature, silhouette, and accessory accents with the event’s specific ambiance, lighting conditions, and seasonal atmosphere."
+    ),
+    "fashion_editor": (
+        "You are the Editor-in-Chief of a prestigious fashion publication. Advise a {gender} trendsetter with a {body_type} frame and {face_shape} face attending the {occasion}, "
+        "currently in {items}. Offer one magazine-cover-worthy styling tip—highlight a trending color palette, editorial-worthy silhouette, and innovative accessory placement that will resonate with a discerning audience."
+    ),
+    "influencer_style": (
+        "As a cutting-edge style influencer with millions of followers, recommend one eye-catching flair tip for a {gender} follower with a {body_type} physique and {face_shape} face, "
+        "heading to the {occasion} in {items}. Frame it as a social-media-caption-ready moment: mention a statement accessory, bold color pop, or texture twist that will go viral."
+    ),
+    "seasonal_trend": (
+        "As a seasonal style expert specializing in spring/summer trends, guide a {gender} individual with a {body_type} shape and {face_shape} face preparing for the {occasion}. "
+        "They currently wear {items}. Provide one tip incorporating current seasonal motifs—think floral prints, breathable linens, or eco-friendly fabrics—that elevates their ensemble."
+    ),
+}
+
+
+# Test parameters
+BODY_TYPE = "Slim"
+FACE_SHAPE = "SQUARE"
+OCCASION = "BEACH PARTY"
+ITEMS = ["jeans", "jacket", "shoes",'shirt']
+GENDER = "Male"
+
+if __name__ == "__main__":
+    advisor = StyleSavvy()
+
+    for name, template in PROMPTS.items():
+        print(f"=== Testing template: {name} ===")
+
+        # Build prompt
+        prompt = template.format(
+            body_type=BODY_TYPE,
+            face_shape=FACE_SHAPE,
+            occasion=OCCASION,
+            gender = GENDER,
+            items=", ".join(ITEMS)
+
+        )
+        print("Prompt:\n" + prompt)
+
+        # Generate response
+        result = advisor.pipe(
+            prompt,
+            max_length=advisor.max_length,
+            early_stopping=True,
+            num_beams=4,
+            no_repeat_ngram_size=3,  
+            do_sample=False)[0]["generated_text"].strip()
+
+        print("\nRaw generated output:\n" + result)
+
+        # Parse into tips (bullets or sentence)
+        lines = result.splitlines()
+        tips = [ln.strip("-*0123456789. ").strip() for ln in lines if ln.strip()]
+        if len(tips) < 3:
+            # fallback to sentence split
+            tips = [p.strip() for p in result.split(".") if p.strip()]
+        tips = list(dict.fromkeys(tips))  # remove duplicates
+
+        print(f"\n💡 Parsed {len(tips)} style tips:")
+        for i, tip in enumerate(tips[:5], 1):
+            print(f"{i}. {tip}")
+        print("-" * 40)