app.py

import os
try:
    import perspective2d
except:
    os.system(f"pip install git+https://github.com/jinlinyi/PerspectiveFields.git@hf-debug")


import gradio as gr
import cv2
import copy
import numpy as np
import os.path as osp
from datetime import datetime

import torch
from PIL import Image, ImageDraw
from glob import glob

from perspective2d import PerspectiveFields
from perspective2d.utils import draw_perspective_fields, draw_from_r_p_f_cx_cy
from perspective2d.perspectivefields import model_zoo




title = "Perspective Fields Demo"

description = """
<p style="text-align: center">
    <a href="https://jinlinyi.github.io/PerspectiveFields/" target="_blank">Project Page</a> | 
    <a href="https://arxiv.org/abs/2212.03239" target="_blank">Paper</a> | 
    <a href="https://github.com/jinlinyi/PerspectiveFields" target="_blank">Code</a> | 
    <a href="https://www.youtube.com/watch?v=sN5B_ZvMva8&themeRefresh=1" target="_blank">Video</a>
</p>
<h2>Gradio Demo</h2>
<p>Try our Gradio demo for Perspective Fields for single image camera calibration. You can click on one of the provided examples or upload your own image.</p>
<h3>Available Models:</h3>
<ol>
    <li><span style="color:red;">[NEW!!!]</span><strong>Paramnet-360Cities-edina:</strong> Our latest model trained on <a href="https://www.360cities.net/">360cities</a> and <a href="https://github.com/tien-d/EgoDepthNormal/tree/main#egocentric-depth-on-everyday-indoor-activities-edina-dataset">EDINA</a> dataset.</li>
    <li><strong>PersNet-360Cities:</strong> PerspectiveNet trained on the 360Cities dataset. This model predicts perspective fields and is designed to be robust and generalize well to both indoor and outdoor images.</li>
    <li><strong>PersNet_Paramnet-GSV-uncentered:</strong> A combination of PerspectiveNet and ParamNet trained on the Google Street View (GSV) dataset. This model predicts camera Roll, Pitch, and Field of View (FoV), as well as the Principal Point location.</li>
    <li><strong>PersNet_Paramnet-GSV-centered:</strong> PerspectiveNet+ParamNet trained on the GSV dataset. This model assumes the principal point is at the center of the image and predicts camera Roll, Pitch, and FoV.</li>
</ol>
"""


article = """
<p style='text-align: center'><a href='https://arxiv.org/abs/2212.03239' target='_blank'>Perspective Fields for Single Image Camera Calibrations</a> | <a href='https://github.com/jinlinyi/PerspectiveFields' target='_blank'>Github Repo</a></p>
"""



def resize_fix_aspect_ratio(img, field, target_width=None, target_height=None):
    height = img.shape[0]
    width = img.shape[1]
    if target_height is None:
        factor = target_width / width
    elif target_width is None:
        factor = target_height / height
    else:
        factor = max(target_width / width, target_height / height)
    if factor == target_width / width:
        target_height = int(height * factor)
    else:
        target_width = int(width * factor)
        
    img = cv2.resize(img, (target_width, target_height))
    for key in field:
        if key not in ['up', 'lati']:
            continue
        tmp = field[key].numpy()
        transpose = len(tmp.shape) == 3
        if transpose:
            tmp = tmp.transpose(1,2,0)
        tmp = cv2.resize(tmp, (target_width, target_height))
        if transpose:
            tmp = tmp.transpose(2,0,1)
        field[key] = torch.tensor(tmp)
    return img, field


def inference(img_rgb, model_type):
    if model_type is None:
        return None, ""
    pf_model = PerspectiveFields(model_type).eval().to(device)
    pred = pf_model.inference(img_bgr=img_rgb[...,::-1])
    img_h = img_rgb.shape[0]
    field = {
        'up': pred['pred_gravity_original'].cpu().detach(),
        'lati': pred['pred_latitude_original'].cpu().detach(),
    }
    img_rgb, field = resize_fix_aspect_ratio(img_rgb, field, 640)
    if not model_zoo[model_type]['param']:
        pred_vis = draw_perspective_fields(
            img_rgb,
            field['up'],
            torch.deg2rad(field['lati']),
            color=(0,1,0),
        )
        param = "Not Implemented"
    else:
        r_p_f_rad = np.radians(
            [
                pred['pred_roll'].cpu().item(),
                pred['pred_pitch'].cpu().item(),
                pred['pred_general_vfov'].cpu().item(),
            ]
        )
        cx_cy = [
            pred['pred_rel_cx'].cpu().item(),
            pred['pred_rel_cy'].cpu().item(),
        ]
        param = f"roll {pred['pred_roll'].cpu().item() :.2f}\npitch {pred['pred_pitch'].cpu().item() :.2f}\nvertical fov {pred['pred_general_vfov'].cpu().item() :.2f}\nfocal_length {pred['pred_rel_focal'].cpu().item()*img_h :.2f}\n"
        param += f"principal point {pred['pred_rel_cx'].cpu().item() :.2f} {pred['pred_rel_cy'].cpu().item() :.2f}"
        pred_vis = draw_from_r_p_f_cx_cy(
            img_rgb, 
            *r_p_f_rad,
            *cx_cy,
            'rad',
            up_color=(0,1,0),
        )
    print(f"""time {datetime.now().strftime("%H:%M:%S")}
    img.shape {img_rgb.shape}
    model_type {model_type}
    param {param}
    """
    )
    return Image.fromarray(pred_vis), param

examples = []
for img_name in glob('assets/imgs/*.*g'):
    examples.append([img_name])
print(examples)

device = 'cuda' if torch.cuda.is_available() else 'cpu'

info = """Select model\n"""
gr.Interface(
    fn=inference,
    inputs=[
        "image", 
        gr.Radio(
            list(model_zoo.keys()), 
            value=list(sorted(model_zoo.keys()))[0], 
            label="Model", 
            info=info,
        ),
    ],
    outputs=[gr.Image(label='Perspective Fields'), gr.Textbox(label='Pred Camera Parameters')],
    title=title,
    description=description,
    article=article,
    examples=examples,
).launch()