Real-ESRGAN Upscaling

Upscale and enhance images using Real-ESRGAN on GPU.

All examples can be run on GPU servers rented through CLORE.AI Marketplace.

Renting on CLORE.AI

1. Visit CLORE.AI Marketplace

2. Filter by GPU type, VRAM, and price

3. Choose On-Demand (fixed rate) or Spot (bid price)

4. Configure your order:

   • Select Docker image

   • Set ports (TCP for SSH, HTTP for web UIs)

   • Add environment variables if needed

   • Enter startup command

5. Select payment: CLORE, BTC, or USDT/USDC

6. Create order and wait for deployment

Access Your Server

• Find connection details in My Orders

• Web interfaces: Use the HTTP port URL

• SSH: ssh -p <port> root@<proxy-address>

What is Real-ESRGAN?

Real-ESRGAN is a practical image restoration model that:

• Upscales images 2x-4x

• Removes noise and artifacts

• Enhances details

• Works on photos, anime, and art

Model Variants

Model	Best For	Speed
RealESRGAN_x4plus	General photos	Medium
RealESRGAN_x4plus_anime_6B	Anime/drawings	Medium
RealESRGAN_x2plus	2x upscale	Fast
RealESRNet_x4plus	Fast processing	Fastest

Quick Deploy

Docker Image:

pytorch/pytorch:2.5.1-cuda12.4-cudnn9-runtime

Ports:

22/tcp
7860/http

Command:

pip install realesrgan gradio && \
python -c "
import gradio as gr
from realesrgan import RealESRGANer
from basicsr.archs.rrdbnet_arch import RRDBNet
import numpy as np
from PIL import Image
import torch

# Load model
model = RRDBNet(num_in_ch=3, num_out_ch=3, num_feat=64, num_block=23, num_grow_ch=32, scale=4)
upsampler = 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=True
)

def upscale(image, scale):
    img = np.array(image)
    output, _ = upsampler.enhance(img, outscale=scale)
    return Image.fromarray(output)

demo = gr.Interface(
    fn=upscale,
    inputs=[gr.Image(type='pil'), gr.Slider(2, 4, value=4, step=1, label='Scale')],
    outputs=gr.Image(type='pil'),
    title='Real-ESRGAN Upscaler'
)
demo.launch(server_name='0.0.0.0', server_port=7860)
"

Accessing Your Service

After deployment, find your http_pub URL in My Orders:

1. Go to My Orders page

2. Click on your order

3. Find the http_pub URL (e.g., abc123.clorecloud.net)

Use https://YOUR_HTTP_PUB_URL instead of localhost in examples below.

CLI Usage

Installation

pip install realesrgan

Basic Upscaling

# Upscale single image
python -m realesrgan -i input.jpg -o output.jpg -n RealESRGAN_x4plus -s 4

# Upscale folder
python -m realesrgan -i ./inputs -o ./outputs -n RealESRGAN_x4plus -s 4

Options

python -m realesrgan \
    -i input.jpg \           # Input
    -o output.jpg \          # Output
    -n RealESRGAN_x4plus \   # Model name
    -s 4 \                   # Scale factor
    --face_enhance \         # Enable face enhancement
    --fp32 \                 # Use FP32 (more VRAM, better quality)
    --tile 400               # Tile size for large images

Python API

Basic Usage

from realesrgan import RealESRGANer
from basicsr.archs.rrdbnet_arch import RRDBNet
import cv2

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

upsampler = RealESRGANer(
    scale=4,
    model_path='RealESRGAN_x4plus.pth',
    model=model,
    tile=0,
    tile_pad=10,
    pre_pad=0,
    half=True  # Use FP16
)

# Upscale
img = cv2.imread('input.jpg', cv2.IMREAD_UNCHANGED)
output, _ = upsampler.enhance(img, outscale=4)
cv2.imwrite('output.jpg', output)

With Face Enhancement

from realesrgan import RealESRGANer
from basicsr.archs.rrdbnet_arch import RRDBNet
from gfpgan import GFPGANer
import cv2

# Real-ESRGAN model
model = RRDBNet(num_in_ch=3, num_out_ch=3, num_feat=64, num_block=23, num_grow_ch=32, scale=4)
upsampler = RealESRGANer(scale=4, model_path='RealESRGAN_x4plus.pth', model=model, half=True)

# GFPGAN for faces
face_enhancer = GFPGANer(
    model_path='GFPGANv1.4.pth',
    upscale=4,
    arch='clean',
    channel_multiplier=2,
    bg_upsampler=upsampler
)

# Process
img = cv2.imread('portrait.jpg')
_, _, output = face_enhancer.enhance(img, has_aligned=False, only_center_face=False, paste_back=True)
cv2.imwrite('output.jpg', output)

Anime Upscaling

from realesrgan import RealESRGANer
from basicsr.archs.rrdbnet_arch import RRDBNet
import cv2

# Anime model
model = RRDBNet(num_in_ch=3, num_out_ch=3, num_feat=64, num_block=6, num_grow_ch=32, scale=4)

upsampler = RealESRGANer(
    scale=4,
    model_path='RealESRGAN_x4plus_anime_6B.pth',
    model=model,
    half=True
)

img = cv2.imread('anime.png')
output, _ = upsampler.enhance(img, outscale=4)
cv2.imwrite('anime_upscaled.png', output)

Batch Processing

Process Folder

import os
from realesrgan import RealESRGANer
from basicsr.archs.rrdbnet_arch import RRDBNet
import cv2
from tqdm import tqdm

# Setup
model = RRDBNet(num_in_ch=3, num_out_ch=3, num_feat=64, num_block=23, num_grow_ch=32, scale=4)
upsampler = RealESRGANer(scale=4, model_path='RealESRGAN_x4plus.pth', model=model, half=True)

input_dir = './inputs'
output_dir = './outputs'
os.makedirs(output_dir, exist_ok=True)

# Process all images
for filename in tqdm(os.listdir(input_dir)):
    if filename.lower().endswith(('.png', '.jpg', '.jpeg', '.webp')):
        img_path = os.path.join(input_dir, filename)
        img = cv2.imread(img_path, cv2.IMREAD_UNCHANGED)

        output, _ = upsampler.enhance(img, outscale=4)

        output_path = os.path.join(output_dir, f"upscaled_{filename}")
        cv2.imwrite(output_path, output)

Shell Script

#!/bin/bash
INPUT_DIR=$1
OUTPUT_DIR=$2

mkdir -p $OUTPUT_DIR

for file in $INPUT_DIR/*.{jpg,png,jpeg}; do
    if [ -f "$file" ]; then
        filename=$(basename "$file")
        python -m realesrgan -i "$file" -o "$OUTPUT_DIR/upscaled_$filename" -n RealESRGAN_x4plus -s 4
        echo "Processed: $filename"
    fi
done

Tiled Processing (Large Images)

For images that don't fit in VRAM:

upsampler = RealESRGANer(
    scale=4,
    model_path='RealESRGAN_x4plus.pth',
    model=model,
    tile=400,      # Process in 400px tiles
    tile_pad=10,   # Overlap between tiles
    pre_pad=0,
    half=True
)

Tile Size Recommendations

VRAM	Max Tile Size
4GB	200
6GB	300
8GB	400
12GB	600
24GB	0 (no tiling)

Video Upscaling

Using Real-ESRGAN

import cv2
from realesrgan import RealESRGANer
from basicsr.archs.rrdbnet_arch import RRDBNet
from tqdm import tqdm

# Setup model
model = RRDBNet(num_in_ch=3, num_out_ch=3, num_feat=64, num_block=23, num_grow_ch=32, scale=4)
upsampler = RealESRGANer(scale=4, model_path='RealESRGAN_x4plus.pth', model=model, tile=400, half=True)

# Open video
cap = cv2.VideoCapture('input.mp4')
fps = cap.get(cv2.CAP_PROP_FPS)
width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH)) * 4
height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT)) * 4
total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))

# Output
fourcc = cv2.VideoWriter_fourcc(*'mp4v')
out = cv2.VideoWriter('output.mp4', fourcc, fps, (width, height))

# Process frames
for _ in tqdm(range(total_frames)):
    ret, frame = cap.read()
    if not ret:
        break

    output, _ = upsampler.enhance(frame, outscale=4)
    out.write(output)

cap.release()
out.release()

FFmpeg Pipeline

# Extract frames
ffmpeg -i input.mp4 -qscale:v 1 frames/frame_%06d.png

# Upscale frames
python -m realesrgan -i frames -o upscaled -n RealESRGAN_x4plus -s 4

# Reassemble video
ffmpeg -framerate 30 -i upscaled/frame_%06d.png -c:v libx264 -pix_fmt yuv420p output.mp4

# Add audio back
ffmpeg -i output.mp4 -i input.mp4 -c:v copy -c:a aac -map 0:v:0 -map 1:a:0 final.mp4

API Server

FastAPI Server

from fastapi import FastAPI, UploadFile
from fastapi.responses import Response
from realesrgan import RealESRGANer
from basicsr.archs.rrdbnet_arch import RRDBNet
import cv2
import numpy as np
import io

app = FastAPI()

# Load model
model = RRDBNet(num_in_ch=3, num_out_ch=3, num_feat=64, num_block=23, num_grow_ch=32, scale=4)
upsampler = RealESRGANer(scale=4, model_path='RealESRGAN_x4plus.pth', model=model, tile=400, half=True)

@app.post("/upscale")
async def upscale(file: UploadFile, scale: int = 4):
    contents = await file.read()
    nparr = np.frombuffer(contents, np.uint8)
    img = cv2.imdecode(nparr, cv2.IMREAD_UNCHANGED)

    output, _ = upsampler.enhance(img, outscale=scale)

    _, encoded = cv2.imencode('.png', output)
    return Response(content=encoded.tobytes(), media_type="image/png")

# Run: uvicorn server:app --host 0.0.0.0 --port 8000

Usage

curl -X POST "http://localhost:8000/upscale?scale=4" \
    -F "[email protected]" \
    --output upscaled.png

Model Comparison

Model	Quality	Speed	VRAM	Best For
x4plus	Best	Slow	4GB+	Photos
x4plus_anime_6B	Best	Medium	3GB+	Anime
x2plus	Good	Fast	2GB+	Quick 2x
RealESRNet	OK	Fastest	2GB+	Previews

Performance

Image Size	GPU	4x Upscale Time
512x512	RTX 3090	~0.5s
1024x1024	RTX 3090	~1.5s
2048x2048	RTX 3090	~5s
512x512	RTX 4090	~0.3s

Troubleshooting

CUDA Out of Memory

# Use tiling
upsampler = RealESRGANer(..., tile=200, ...)

# Or reduce scale
output, _ = upsampler.enhance(img, outscale=2)  # Instead of 4

Artifacts in Output

• Use smaller tile size with more overlap

• Try different model (anime vs photo)

• Check input image quality

Slow Processing

• Enable FP16: half=True

• Increase tile size if VRAM allows

• Use faster model: RealESRNet

Download Results

scp -P <port> -r root@<proxy>:/workspace/outputs/ ./upscaled_images/

Cost Estimate

Typical CLORE.AI marketplace rates (as of 2024):

GPU	Hourly Rate	Daily Rate	4-Hour Session
RTX 3060	~$0.03	~$0.70	~$0.12
RTX 3090	~$0.06	~$1.50	~$0.25
RTX 4090	~$0.10	~$2.30	~$0.40
A100 40GB	~$0.17	~$4.00	~$0.70
A100 80GB	~$0.25	~$6.00	~$1.00

Prices vary by provider and demand. Check CLORE.AI Marketplace for current rates.

Save money:

• Use Spot market for flexible workloads (often 30-50% cheaper)

• Pay with CLORE tokens

• Compare prices across different providers

Next Steps

• GFPGAN Face Restoration

• AI Video Generation

• Stable Diffusion WebUI