app.py

from fastapi import FastAPI, UploadFile, File, Form
import uuid, os
from PIL import Image

from rembg import remove
from realesrgan import RealESRGANer
from basicsr.archs.rrdbnet_arch import RRDBNet

from gfpgan import GFPGANer
from diffusers import StableDiffusionImg2ImgPipeline
import torch

app = FastAPI()

UPLOAD_DIR = "uploads"
OUTPUT_DIR = "outputs"
os.makedirs(UPLOAD_DIR, exist_ok=True)
os.makedirs(OUTPUT_DIR, exist_ok=True)

DEVICE = "cuda" if torch.cuda.is_available() else "cpu"

# ===== Load models ONCE =====

# Real-ESRGAN
# ===== Real-ESRGAN =====
model = RRDBNet(
    num_in_ch=3,
    num_out_ch=3,
    num_feat=64,
    num_block=23,
    num_grow_ch=32,
    scale=4
)

upscaler = RealESRGANer(
    scale=4,
    model_path="https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth",
    model=model,
    tile=0,
    tile_pad=10,
    pre_pad=0,
    half=(DEVICE == "cuda"),
    device=DEVICE
)

#upscaler.load_weights("RealESRGAN_x4.pth")

# GFPGAN
gfpgan = GFPGANer(
    model_path="https://github.com/TencentARC/GFPGAN/releases/download/v1.3.4/GFPGANv1.4.pth",
    upscale=1,
    arch="clean",
    channel_multiplier=2,
    device=DEVICE
)



# Stable Diffusion Img2Img
dtype = torch.float16 if DEVICE == "cuda" else torch.float32

sd_pipe = StableDiffusionImg2ImgPipeline.from_pretrained(
    "runwayml/stable-diffusion-v1-5",
    torch_dtype=dtype
).to(DEVICE)


# ===== Endpoints =====

@app.post("/remove-bg")
async def remove_bg(file: UploadFile = File(...)):
    img = Image.open(file.file).convert("RGBA")
    out = remove(img)

    fname = f"{uuid.uuid4().hex}.png"
    path = os.path.join(OUTPUT_DIR, fname)
    out.save(path)

    return {"file": fname}


@app.post("/enhance")
async def enhance(file: UploadFile = File(...)):
    img = Image.open(file.file).convert("RGB")
    out, _ = upscaler.enhance(img)

    fname = f"{uuid.uuid4().hex}.png"
    out.save(os.path.join(OUTPUT_DIR, fname))
    return {"file": fname}


@app.post("/upscale")
async def upscale(file: UploadFile = File(...), scale: int = Form(2)):
    img = Image.open(file.file).convert("RGB")
    upscaler.scale = scale
    out, _ = upscaler.enhance(img, outscale=scale)

    fname = f"{uuid.uuid4().hex}.png"
    out.save(os.path.join(OUTPUT_DIR, fname))
    return {"file": fname}


@app.post("/retouch")
async def retouch(file: UploadFile = File(...)):
    img = Image.open(file.file).convert("RGB")
    _, _, out = gfpgan.enhance(
        img,
        has_aligned=False,
        only_center_face=False,
        paste_back=True
    )

    fname = f"{uuid.uuid4().hex}.png"
    Image.fromarray(out).save(os.path.join(OUTPUT_DIR, fname))
    return {"file": fname}


@app.post("/edit")
async def edit_image(
    file: UploadFile = File(...),
    prompt: str = Form(...),
    strength: float = Form(0.6)
):
    img = Image.open(file.file).convert("RGB").resize((512, 512))

    result = sd_pipe(
        prompt=prompt,
        image=img,
        strength=strength,
        guidance_scale=7.5
    ).images[0]

    fname = f"{uuid.uuid4().hex}.png"
    result.save(os.path.join(OUTPUT_DIR, fname))
    return {
        "prompt": prompt,
        "file": fname
    }