update: ci

.flake8

@@ -0,0 +1,4 @@
+[flake8]
+max-line-length = 80
+extend-ignore = E203,E501,E402
+exclude = .git,__pycache__,build,.venv/,third_party

.github/.stale.yml

@@ -0,0 +1,17 @@
+# Number of days of inactivity before an issue becomes stale
+daysUntilStale: 60
+# Number of days of inactivity before a stale issue is closed
+daysUntilClose: 7
+# Issues with these labels will never be considered stale
+exemptLabels:
+  - pinned
+  - security
+# Label to use when marking an issue as stale
+staleLabel: wontfix
+# Comment to post when marking an issue as stale. Set to `false` to disable
+markComment: >
+  This issue has been automatically marked as stale because it has not had
+  recent activity. It will be closed if no further activity occurs. Thank you
+  for your contributions.
+# Comment to post when closing a stale issue. Set to `false` to disable
+closeComment: false

.github/ISSUE_TEMPLATE/bug_report.md

@@ -0,0 +1,30 @@
+---
+name: 🐛 Bug report
+about: If something isn't working 🔧
+title: ""
+labels: bug
+assignees:
+---
+
+## 🐛 Bug Report
+
+<!-- A clear and concise description of what the bug is. -->
+
+## 🔬 How To Reproduce
+
+Steps to reproduce the behavior:
+
+1. ...
+
+### Environment
+
+- OS: [e.g. Linux / Windows / macOS]
+- Python version, get it with:
+
+```bash
+python --version
+```
+
+## 📎 Additional context
+
+<!-- Add any other context about the problem here. -->

.github/ISSUE_TEMPLATE/config.yml

@@ -0,0 +1,3 @@
+# Configuration: https://help.github.com/en/github/building-a-strong-community/configuring-issue-templates-for-your-repository
+
+blank_issues_enabled: false

.github/ISSUE_TEMPLATE/feature_request.md

@@ -0,0 +1,15 @@
+---
+name: 🚀 Feature request
+about: Suggest an idea for this project 🏖
+title: ""
+labels: enhancement
+assignees:
+---
+
+## 🚀 Feature Request
+
+<!-- A clear and concise description of the feature proposal. -->
+
+## 📎 Additional context
+
+<!-- Add any other context or screenshots about the feature request here. -->

.github/ISSUE_TEMPLATE/question.md

@@ -0,0 +1,25 @@
+---
+name: ❓ Question
+about: Ask a question about this project 🎓
+title: ""
+labels: question
+assignees:
+---
+
+## Checklist
+
+<!-- Mark with an `x` all the checkboxes that apply (like `[x]`) -->
+
+- [ ] I've searched the project's [`issues`]
+
+## ❓ Question
+
+<!-- What is your question -->
+
+How can I [...]?
+
+Is it possible to [...]?
+
+## 📎 Additional context
+
+<!-- Add any other context or screenshots about the feature request here. -->

.github/PULL_REQUEST_TEMPLATE.md

@@ -0,0 +1,7 @@
+## Description
+
+<!-- Add a more detailed description of the changes if needed. -->
+
+## Related Issue
+
+<!-- If your PR refers to a related issue, link it here. -->

.github/release-drafter.yml

@@ -0,0 +1,24 @@
+# Release drafter configuration https://github.com/release-drafter/release-drafter#configuration
+# Emojis were chosen to match the https://gitmoji.carloscuesta.me/
+
+name-template: "v$RESOLVED_VERSION"
+tag-template: "v$RESOLVED_VERSION"
+
+categories:
+  - title: ":rocket: Features"
+    labels: [enhancement, feature]
+  - title: ":wrench: Fixes"
+    labels: [bug, bugfix, fix]
+  - title: ":toolbox: Maintenance & Refactor"
+    labels: [refactor, refactoring, chore]
+  - title: ":package: Build System & CI/CD & Test"
+    labels: [build, ci, testing, test]
+  - title: ":pencil: Documentation"
+    labels: [documentation]
+  - title: ":arrow_up: Dependencies updates"
+    labels: [dependencies]
+
+template: |
+  ## What’s Changed
+
+  $CHANGES

.github/workflows/ci.yml

@@ -0,0 +1,32 @@
+name: CI CPU
+
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+    branches:
+      - main
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+
+    steps:
+    - name: Checkout code
+      uses: actions/checkout@v4
+      with:
+        submodules: recursive
+
+    - name: Set up Python
+      uses: actions/setup-python@v4
+      with:
+        python-version: "3.10"
+
+    - name: Install dependencies
+      run: |
+        pip install -r requirements.txt
+        sudo apt-get update && sudo apt-get install ffmpeg libsm6 libxext6 -y
+
+    - name: Run tests
+      run: python test_app_cli.py

.github/workflows/format.yml

@@ -0,0 +1,24 @@
+name: Format and Lint Checks
+on:
+  push:
+    branches:
+      - main
+    paths:
+      - '*.py'
+  pull_request:
+    types: [ assigned, opened, synchronize, reopened ]
+jobs:
+  check:
+    name: Format and Lint Checks
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v4
+        with:
+          python-version: '3.10'
+          cache: 'pip'
+      - run: python -m pip install --upgrade pip
+      - run: python -m pip install .[dev]
+      - run: python -m flake8 common/*.py hloc/*.py hloc/matchers/*.py hloc/extractors/*.py 
+      - run: python -m isort common/*.py hloc/*.py hloc/matchers/*.py hloc/extractors/*.py --check-only --diff
+      - run: python -m black common/*.py hloc/*.py hloc/matchers/*.py hloc/extractors/*.py --check --diff

.github/workflows/release-drafter.yml

@@ -0,0 +1,16 @@
+name: Release Drafter
+
+on:
+  push:
+    # branches to consider in the event; optional, defaults to all
+    branches:
+      - master
+
+jobs:
+  update_release_draft:
+    runs-on: ubuntu-latest
+    steps:
+      # Drafts your next Release notes as Pull Requests are merged into "master"
+      - uses: release-drafter/[email protected]
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.gitignore

@@ -14,9 +14,11 @@ experiments
 third_party/REKD
 hloc/matchers/dedode.py
 gradio_cached_examples
-
+*.mp4
 hloc/matchers/quadtree.py
 third_party/QuadTreeAttention
 desktop.ini
+*.egg-info
+output.pkl
 experiments*
 gen_example.py

README.md

@@ -44,6 +44,8 @@ The tool currently supports various popular image matching algorithms, namely:
 - [ ] [DUSt3R](https://github.com/naver/dust3r), arXiv 2023
 - [x] [LightGlue](https://github.com/cvg/LightGlue), ICCV 2023
 - [x] [DarkFeat](https://github.com/THU-LYJ-Lab/DarkFeat), AAAI 2023
+- [x] [SFD2](https://github.com/feixue94/sfd2), CVPR 2023
+- [x] [IMP](https://github.com/feixue94/imp-release), CVPR 2023
 - [ ] [ASTR](https://github.com/ASTR2023/ASTR), CVPR 2023
 - [ ] [SEM](https://github.com/SEM2023/SEM), CVPR 2023
 - [ ] [DeepLSD](https://github.com/cvg/DeepLSD), CVPR 2023

common/api.py

@@ -1,26 +1,23 @@
-import cv2
-import torch
 import warnings
-import numpy as np
 from pathlib import Path
-from typing import Dict, Any, Optional, Tuple, List, Union
-from hloc import logger
-from hloc import match_dense, match_features, extract_features
+from typing import Any, Dict, Optional
+
+import cv2
+import matplotlib.pyplot as plt
+import numpy as np
+import torch
+
+from hloc import extract_features, logger, match_dense, match_features
 from hloc.utils.viz import add_text, plot_keypoints
+
 from .utils import (
-    load_config,
-    get_model,
-    get_feature_model,
-    filter_matches,
-    device,
     ROOT,
+    filter_matches,
+    get_feature_model,
+    get_model,
+    load_config,
 )
-from .viz import (
-    fig2im,
-    plot_images,
-    display_matches,
-)
-import matplotlib.pyplot as plt
+from .viz import display_matches, fig2im, plot_images
 
 warnings.simplefilter("ignore")
 
@@ -109,7 +106,7 @@ class ImageMatchingAPI(torch.nn.Module):
                     "match_threshold"
                 ] = match_threshold
             except TypeError as e:
-                breakpoint()
+                logger.error(e)
         else:
             self.conf["feature"]["model"]["max_keypoints"] = max_keypoints
             self.conf["feature"]["model"][
@@ -137,7 +134,9 @@ class ImageMatchingAPI(torch.nn.Module):
                 self.match_conf["preprocessing"],
                 device=self.device,
             )
-            last_fixed = "{}".format(self.match_conf["model"]["name"])
+            last_fixed = "{}".format(  # noqa: F841
+                self.match_conf["model"]["name"]
+            )
         else:
             pred0 = extract_features.extract(
                 self.extractor, img0, self.extract_conf["preprocessing"]
@@ -290,7 +289,5 @@ class ImageMatchingAPI(torch.nn.Module):
 
 
 if __name__ == "__main__":
-    import argparse
-
     config = load_config(ROOT / "common/config.yaml")
-    test_api(config)
+    api = ImageMatchingAPI(config)

common/app_class.py

@@ -1,22 +1,21 @@
-import argparse
-import numpy as np
-import gradio as gr
 from pathlib import Path
-from typing import Dict, Any, Optional, Tuple, List, Union
+from typing import Any, Dict, Optional, Tuple
+
+import gradio as gr
+import numpy as np
+
 from common.utils import (
-    ransac_zoo,
+    GRADIO_VERSION,
+    gen_examples,
     generate_warp_images,
-    load_config,
     get_matcher_zoo,
+    load_config,
+    ransac_zoo,
     run_matching,
     run_ransac,
     send_to_match,
-    gen_examples,
-    GRADIO_VERSION,
-    ROOT,
 )
 
-
 DESCRIPTION = """
 # Image Matching WebUI
 This Space demonstrates [Image Matching WebUI](https://github.com/Vincentqyw/image-matching-webui) by vincent qin. Feel free to play with it, or duplicate to run image matching without a queue!
@@ -132,12 +131,14 @@ class ImageMatchingApp:
                                         label="Keypoint thres.",
                                         value=0.015,
                                     )
-                                    detect_line_threshold = gr.Slider(
-                                        minimum=0.1,
-                                        maximum=1,
-                                        step=0.01,
-                                        label="Line thres.",
-                                        value=0.2,
+                                    detect_line_threshold = (  # noqa: F841
+                                        gr.Slider(
+                                            minimum=0.1,
+                                            maximum=1,
+                                            step=0.01,
+                                            label="Line thres.",
+                                            value=0.2,
+                                        )
                                     )
                                 # matcher_lists = gr.Radio(
                                 #     ["NN-mutual", "Dual-Softmax"],

common/config.yaml

@@ -30,20 +30,22 @@ matcher_zoo:
   DUSt3R:
     # TODO: duster is under development  
     enable: true
+    skip_ci: true
     matcher: duster
     dense: true
-    info: 
+    info:
       name: DUSt3R #dispaly name
       source: "CVPR 2024"
       github: https://github.com/naver/dust3r
       paper: https://arxiv.org/abs/2312.14132
       project: https://dust3r.europe.naverlabs.com
-      display: true  
+      display: true
   GIM(dkm):
     enable: true
+    skip_ci: true
     matcher: gim(dkm)
     dense: true
-    info: 
+    info:
       name: GIM(DKM) #dispaly name
       source: "ICLR 2024"
       github: https://github.com/xuelunshen/gim
@@ -52,8 +54,9 @@ matcher_zoo:
       display: true
   RoMa:
     matcher: roma
+    skip_ci: true
     dense: true
-    info: 
+    info:
       name: RoMa #dispaly name
       source: "CVPR 2024"
       github: https://github.com/Parskatt/RoMa
@@ -62,8 +65,9 @@ matcher_zoo:
       display: true
   dkm:
     matcher: dkm
+    skip_ci: true
     dense: true
-    info: 
+    info:
       name: DKM #dispaly name
       source: "CVPR 2023"
       github: https://github.com/Parskatt/DKM
@@ -73,7 +77,7 @@ matcher_zoo:
   loftr:
     matcher: loftr
     dense: true
-    info: 
+    info:
       name: LoFTR #dispaly name
       source: "CVPR 2021"
       github: https://github.com/zju3dv/LoFTR
@@ -82,6 +86,7 @@ matcher_zoo:
       display: true
   cotr:
     enable: false
+    skip_ci: true
     matcher: cotr
     dense: true
     info:
@@ -363,3 +368,31 @@ matcher_zoo:
       paper: https://arxiv.org/abs/2104.03362
       project: null
       display: true
+
+  sfd2+imp:
+    matcher: imp
+    feature: sfd2
+    enable: false
+    dense: false
+    skip_ci: true
+    info:
+      name: SFD2+IMP #dispaly name
+      source: "CVPR 2023"
+      github: https://github.com/feixue94/imp-release
+      paper: https://arxiv.org/pdf/2304.14837
+      project: https://feixue94.github.io/
+      display: true
+
+  sfd2+mnn:
+    matcher: NN-mutual
+    feature: sfd2
+    enable: false
+    dense: false
+    skip_ci: true
+    info:
+      name: SFD2+MNN #dispaly name
+      source: "CVPR 2023"
+      github: https://github.com/feixue94/sfd2
+      paper: https://arxiv.org/abs/2304.14845
+      project: https://feixue94.github.io/
+      display: true

common/utils.py

@@ -1,35 +1,35 @@
 import os
-import cv2
-import sys
-import torch
+import pickle
 import random
-import psutil
 import shutil
-import numpy as np
-import gradio as gr
-from PIL import Image
+import time
+import warnings
+from itertools import combinations
 from pathlib import Path
+from typing import Any, Callable, Dict, List, Optional, Tuple, Union
+
+import cv2
+import gradio as gr
+import matplotlib.pyplot as plt
+import numpy as np
 import poselib
-from itertools import combinations
-from typing import Callable, Dict, Any, Optional, Tuple, List, Union
-from hloc import matchers, extractors, logger
-from hloc.utils.base_model import dynamic_load
-from hloc import match_dense, match_features, extract_features
-from .viz import (
-    fig2im,
-    plot_images,
-    display_matches,
-    display_keypoints,
-    plot_color_line_matches,
+import psutil
+from PIL import Image
+
+from hloc import (
+    DEVICE,
+    extract_features,
+    extractors,
+    logger,
+    match_dense,
+    match_features,
+    matchers,
 )
-import time
-import matplotlib.pyplot as plt
-import warnings
-import tempfile
-import pickle
+from hloc.utils.base_model import dynamic_load
+
+from .viz import display_keypoints, display_matches, fig2im, plot_images
 
 warnings.simplefilter("ignore")
-device = "cuda" if torch.cuda.is_available() else "cpu"
 
 ROOT = Path(__file__).parent.parent
 # some default values
@@ -91,14 +91,13 @@ class ModelCache:
         host_colocation = int(os.environ.get("HOST_COLOCATION", "1"))
         vm = psutil.virtual_memory()
         du = shutil.disk_usage(".")
-        vm_ratio = host_colocation * vm.used / vm.total
         if verbose:
             logger.info(
                 f"RAM: {vm.used / 1e9:.1f}/{vm.total / host_colocation / 1e9:.1f}GB"
             )
-            # logger.info(
-            #     f"DISK: {du.used / 1e9:.1f}/{du.total / host_colocation / 1e9:.1f}GB"
-            # )
+            logger.info(
+                f"DISK: {du.used / 1e9:.1f}/{du.total / host_colocation / 1e9:.1f}GB"
+            )
         return vm.used / 1e9
 
     def print_memory_usage(self):
@@ -173,7 +172,7 @@ def get_model(match_conf: Dict[str, Any]):
         A matcher model instance.
     """
     Model = dynamic_load(matchers, match_conf["model"]["name"])
-    model = Model(match_conf["model"]).eval().to(device)
+    model = Model(match_conf["model"]).eval().to(DEVICE)
     return model
 
 
@@ -188,7 +187,7 @@ def get_feature_model(conf: Dict[str, Dict[str, Any]]):
         A feature extraction model instance.
     """
     Model = dynamic_load(extractors, conf["model"]["name"])
-    model = Model(conf["model"]).eval().to(device)
+    model = Model(conf["model"]).eval().to(DEVICE)
     return model
 
 
@@ -423,7 +422,7 @@ def _filter_matches_poselib(
     elif geometry_type == "Fundamental":
         M, info = poselib.estimate_fundamental(kp0, kp1, ransac_options)
     else:
-        raise notImplementedError("Not Implemented")
+        raise NotImplementedError
 
     return M, np.array(info["inliers"])
 
@@ -464,7 +463,7 @@ def proc_ransac_matches(
             geometry_type,
         )
     else:
-        raise notImplementedError("Not Implemented")
+        raise NotImplementedError
 
 
 def filter_matches(
@@ -617,7 +616,9 @@ def compute_geometry(
                 geo_info["H1"] = H1.tolist()
                 geo_info["H2"] = H2.tolist()
             except cv2.error as e:
-                logger.error(f"StereoRectifyUncalibrated failed, skip!")
+                logger.error(
+                    f"StereoRectifyUncalibrated failed, skip! error: {e}"
+                )
         return geo_info
     else:
         return {}
@@ -643,7 +644,6 @@ def wrap_images(
     """
     h0, w0, _ = img0.shape
     h1, w1, _ = img1.shape
-    result_matrix: Optional[np.ndarray] = None
     if geo_info is not None and len(geo_info) != 0:
         rectified_image0 = img0
         rectified_image1 = None
@@ -656,7 +656,6 @@ def wrap_images(
         title: List[str] = []
         if geom_type == "Homography":
             rectified_image1 = cv2.warpPerspective(img1, H, (w0, h0))
-            result_matrix = H
             title = ["Image 0", "Image 1 - warped"]
         elif geom_type == "Fundamental":
             if geom_type not in geo_info:
@@ -666,7 +665,6 @@ def wrap_images(
                 H1, H2 = np.array(geo_info["H1"]), np.array(geo_info["H2"])
                 rectified_image0 = cv2.warpPerspective(img0, H1, (w0, h0))
                 rectified_image1 = cv2.warpPerspective(img1, H2, (w1, h1))
-                result_matrix = np.array(geo_info["Fundamental"])
                 title = ["Image 0 - warped", "Image 1 - warped"]
         else:
             print("Error: Unknown geometry type")
@@ -705,7 +703,7 @@ def generate_warp_images(
     ):
         return None, None
     geom_info = matches_info["geom_info"]
-    wrapped_images = None
+    warped_image = None
     if choice != "No":
         wrapped_image_pair, warped_image = wrap_images(
             input_image0, input_image1, geom_info, choice
@@ -805,7 +803,7 @@ def run_ransac(
     with open(tmp_state_cache, "wb") as f:
         pickle.dump(state_cache, f)
 
-    logger.info(f"Dump results done!")
+    logger.info("Dump results done!")
 
     return (
         output_matches_ransac,
@@ -880,7 +878,7 @@ def run_matching(
     output_matches_ransac = None
 
     # super slow!
-    if "roma" in key.lower() and device == "cpu":
+    if "roma" in key.lower() and DEVICE == "cpu":
         gr.Info(
             f"Success! Please be patient and allow for about 2-3 minutes."
             f" Due to CPU inference, {key} is quiet slow."
@@ -905,7 +903,7 @@ def run_matching(
 
     if model["dense"]:
         pred = match_dense.match_images(
-            matcher, image0, image1, match_conf["preprocessing"], device=device
+            matcher, image0, image1, match_conf["preprocessing"], device=DEVICE
         )
         del matcher
         extract_conf = None
@@ -1000,7 +998,7 @@ def run_matching(
     tmp_state_cache = "output.pkl"
     with open(tmp_state_cache, "wb") as f:
         pickle.dump(state_cache, f)
-    logger.info(f"Dump results done!")
+    logger.info("Dump results done!")
     return (
         output_keypoints,
         output_matches_raw,

common/viz.py

@@ -1,11 +1,13 @@
-import cv2
 import typing
+from pathlib import Path
+from typing import Dict, List, Optional, Tuple, Union
+
+import cv2
 import matplotlib
+import matplotlib.pyplot as plt
 import numpy as np
 import seaborn as sns
-import matplotlib.pyplot as plt
-from pathlib import Path
-from typing import Dict, Any, Optional, Tuple, List, Union
+
 from hloc.utils.viz import add_text, plot_keypoints
 
 

docker/build_docker.bat

@@ -0,0 +1,3 @@
+docker build -t image-matching-webui:latest . --no-cache
+# docker tag image-matching-webui:latest vincentqin/image-matching-webui:latest
+# docker push vincentqin/image-matching-webui:latest

docker/run_docker.sh

@@ -0,0 +1 @@
+docker run -it -p 7860:7860 vincentqin/image-matching-webui:latest python app.py --server_name "0.0.0.0" --server_port=7860

env-docker.txt

@@ -1,33 +0,0 @@
-e2cnn==0.2.3
-einops==0.6.1
-gdown==4.7.1
-gradio==4.28.3
-gradio_client==0.16.0
-h5py==3.9.0
-imageio==2.31.1
-Jinja2==3.1.2
-kornia==0.6.12
-loguru==0.7.0
-matplotlib==3.7.1
-numpy==1.23.5
-omegaconf==2.3.0
-opencv-contrib-python==4.6.0.66
-opencv-python==4.6.0.66
-pandas==2.0.3
-plotly==5.15.0
-protobuf==4.23.2
-pycolmap==0.5.0
-pytlsd==0.0.2
-pytorch-lightning==1.4.9
-PyYAML==6.0
-scikit-image==0.21.0
-scikit-learn==1.2.2
-scipy==1.11.1
-seaborn==0.12.2
-shapely==2.0.1
-tensorboardX==2.6.1
-torchmetrics==0.6.0
-torchvision==0.17.1
-tqdm==4.65.0
-yacs==0.1.8
-onnxruntime

format.sh

@@ -0,0 +1,3 @@
+python -m flake8 common/*.py hloc/*.py hloc/matchers/*.py hloc/extractors/*.py 
+python -m isort common/*.py hloc/*.py hloc/matchers/*.py hloc/extractors/*.py
+python -m black common/*.py hloc/*.py hloc/matchers/*.py hloc/extractors/*.py

hloc/__init__.py

@@ -1,4 +1,5 @@
 import logging
+
 import torch
 from packaging import version
 

hloc/extract_features.py

@@ -1,21 +1,22 @@
 import argparse
-import torch
+import collections.abc as collections
+import pprint
 from pathlib import Path
-from typing import Dict, List, Union, Optional
-import h5py
 from types import SimpleNamespace
+from typing import Dict, List, Optional, Union
+
 import cv2
+import h5py
 import numpy as np
-from tqdm import tqdm
-import pprint
-import collections.abc as collections
 import PIL.Image
+import torch
 import torchvision.transforms.functional as F
+from tqdm import tqdm
+
 from . import extractors, logger
 from .utils.base_model import dynamic_load
+from .utils.io import list_h5_names, read_image
 from .utils.parsers import parse_image_lists
-from .utils.io import read_image, list_h5_names
-
 
 """
 A set of standard configurations that can be directly selected from the command
@@ -290,6 +291,23 @@ confs = {
             "dfactor": 8,
         },
     },
+    "sfd2": {
+        "output": "feats-sfd2-n4096-r1600",
+        "model": {
+            "name": "sfd2",
+            "max_keypoints": 4096,
+        },
+        "preprocessing": {
+            "grayscale": False,
+            "force_resize": True,
+            "resize_max": 1600,
+            "width": 640,
+            "height": 480,
+            "conf_th": 0.001,
+            "multiscale": False,
+            "scales": [1.0],
+        },
+    },
     # Global descriptors
     "dir": {
         "output": "global-feats-dir",
@@ -460,7 +478,7 @@ def extract(model, image_0, conf):
     #     image0 = image_0[:, :, ::-1]  # BGR to RGB
     data = preprocess(image0, conf)
     pred = model({"image": data["image"]})
-    pred["image_size"] = original_size = data["original_size"]
+    pred["image_size"] = data["original_size"]
     pred = {**pred, **data}
     return pred
 

hloc/extractors/alike.py

@@ -1,10 +1,12 @@
 import sys
 from pathlib import Path
+
 import torch
 
-from ..utils.base_model import BaseModel
 from hloc import logger
 
+from ..utils.base_model import BaseModel
+
 alike_path = Path(__file__).parent / "../../third_party/ALIKE"
 sys.path.append(str(alike_path))
 from alike import ALike as Alike_
@@ -34,7 +36,7 @@ class Alike(BaseModel):
             scores_th=conf["detection_threshold"],
             n_limit=conf["max_keypoints"],
         )
-        logger.info(f"Load Alike model done.")
+        logger.info("Load Alike model done.")
 
     def _forward(self, data):
         image = data["image"]

hloc/extractors/d2net.py

@@ -1,17 +1,17 @@
+import subprocess
 import sys
 from pathlib import Path
-import subprocess
+
 import torch
 
-from ..utils.base_model import BaseModel
 from hloc import logger
 
-d2net_path = Path(__file__).parent / "../../third_party"
-sys.path.append(str(d2net_path))
-from d2net.lib.model_test import D2Net as _D2Net
-from d2net.lib.pyramid import process_multiscale
+from ..utils.base_model import BaseModel
 
 d2net_path = Path(__file__).parent / "../../third_party/d2net"
+sys.path.append(str(d2net_path))
+from lib.model_test import D2Net as _D2Net
+from lib.pyramid import process_multiscale
 
 
 class D2Net(BaseModel):
@@ -30,6 +30,7 @@ class D2Net(BaseModel):
             model_file.parent.mkdir(exist_ok=True)
             cmd = [
                 "wget",
+                "--quiet",
                 "https://dusmanu.com/files/d2-net/" + conf["model_name"],
                 "-O",
                 str(model_file),
@@ -39,7 +40,7 @@ class D2Net(BaseModel):
         self.net = _D2Net(
             model_file=model_file, use_relu=conf["use_relu"], use_cuda=False
         )
-        logger.info(f"Load D2Net model done.")
+        logger.info("Load D2Net model done.")
 
     def _forward(self, data):
         image = data["image"]

hloc/extractors/darkfeat.py

@@ -1,9 +1,12 @@
 import sys
 from pathlib import Path
+
 import subprocess
-from ..utils.base_model import BaseModel
+
 from hloc import logger
 
+from ..utils.base_model import BaseModel
+
 darkfeat_path = Path(__file__).parent / "../../third_party/DarkFeat"
 sys.path.append(str(darkfeat_path))
 from darkfeat import DarkFeat as DarkFeat_
@@ -43,7 +46,7 @@ class DarkFeat(BaseModel):
                     raise e
 
         self.net = DarkFeat_(model_path)
-        logger.info(f"Load DarkFeat model done.")
+        logger.info("Load DarkFeat model done.")
 
     def _forward(self, data):
         pred = self.net({"image": data["image"]})

hloc/extractors/dedode.py

@@ -1,16 +1,18 @@
+import subprocess
 import sys
 from pathlib import Path
-import subprocess
+
 import torch
-from PIL import Image
-from ..utils.base_model import BaseModel
-from hloc import logger
 import torchvision.transforms as transforms
 
+from hloc import logger
+
+from ..utils.base_model import BaseModel
+
 dedode_path = Path(__file__).parent / "../../third_party/DeDoDe"
 sys.path.append(str(dedode_path))
 
-from DeDoDe import dedode_detector_L, dedode_descriptor_B
+from DeDoDe import dedode_descriptor_B, dedode_detector_L
 from DeDoDe.utils import to_pixel_coords
 
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
@@ -49,14 +51,14 @@ class DeDoDe(BaseModel):
         if not model_detector_path.exists():
             model_detector_path.parent.mkdir(exist_ok=True)
             link = self.weight_urls[conf["model_detector_name"]]
-            cmd = ["wget", link, "-O", str(model_detector_path)]
+            cmd = ["wget", "--quiet", link, "-O", str(model_detector_path)]
             logger.info(f"Downloading the DeDoDe detector model with `{cmd}`.")
             subprocess.run(cmd, check=True)
 
         if not model_descriptor_path.exists():
             model_descriptor_path.parent.mkdir(exist_ok=True)
             link = self.weight_urls[conf["model_descriptor_name"]]
-            cmd = ["wget", link, "-O", str(model_descriptor_path)]
+            cmd = ["wget", "--quiet", link, "-O", str(model_descriptor_path)]
             logger.info(
                 f"Downloading the DeDoDe descriptor model with `{cmd}`."
             )
@@ -73,7 +75,7 @@ class DeDoDe(BaseModel):
         self.descriptor = dedode_descriptor_B(
             weights=weights_descriptor, device=device
         )
-        logger.info(f"Load DeDoDe model done.")
+        logger.info("Load DeDoDe model done.")
 
     def _forward(self, data):
         """

hloc/extractors/dir.py

@@ -1,10 +1,11 @@
+import os
 import sys
 from pathlib import Path
-import torch
 from zipfile import ZipFile
-import os
-import sklearn
+
 import gdown
+import sklearn
+import torch
 
 from ..utils.base_model import BaseModel
 
@@ -13,8 +14,8 @@ sys.path.append(
 )
 os.environ["DB_ROOT"] = ""  # required by dirtorch
 
-from dirtorch.utils import common  # noqa: E402
 from dirtorch.extract_features import load_model  # noqa: E402
+from dirtorch.utils import common  # noqa: E402
 
 # The DIR model checkpoints (pickle files) include sklearn.decomposition.pca,
 # which has been deprecated in sklearn v0.24

hloc/extractors/disk.py

@@ -15,7 +15,7 @@ class DISK(BaseModel):
 
     def _init(self, conf):
         self.model = kornia.feature.DISK.from_pretrained(conf["weights"])
-        logger.info(f"Load DISK model done.")
+        logger.info("Load DISK model done.")
 
     def _forward(self, data):
         image = data["image"]

hloc/extractors/dog.py

@@ -1,15 +1,14 @@
 import kornia
+import numpy as np
+import pycolmap
+import torch
 from kornia.feature.laf import (
-    laf_from_center_scale_ori,
     extract_patches_from_pyramid,
+    laf_from_center_scale_ori,
 )
-import numpy as np
-import torch
-import pycolmap
 
 from ..utils.base_model import BaseModel
 
-
 EPS = 1e-6
 
 

hloc/extractors/example.py

@@ -1,9 +1,9 @@
 import sys
 from pathlib import Path
-import subprocess
+
 import torch
-from .. import logger
 
+from .. import logger
 from ..utils.base_model import BaseModel
 
 example_path = Path(__file__).parent / "../../third_party/example"
@@ -35,7 +35,7 @@ class Example(BaseModel):
         # self.net = ExampleNet(is_test=True)
         state_dict = torch.load(model_path, map_location="cpu")
         self.net.load_state_dict(state_dict["model_state"])
-        logger.info(f"Load example model done.")
+        logger.info("Load example model done.")
 
     def _forward(self, data):
         # data: dict, keys: 'image'

hloc/extractors/fire.py

@@ -1,7 +1,8 @@
-from pathlib import Path
-import subprocess
 import logging
+import subprocess
 import sys
+from pathlib import Path
+
 import torch
 import torchvision.transforms as tvf
 
@@ -42,11 +43,11 @@ class FIRe(BaseModel):
         if not model_path.exists():
             model_path.parent.mkdir(exist_ok=True)
             link = self.fire_models[conf["model_name"]]
-            cmd = ["wget", link, "-O", str(model_path)]
+            cmd = ["wget", "--quiet", link, "-O", str(model_path)]
             logger.info(f"Downloading the FIRe model with `{cmd}`.")
             subprocess.run(cmd, check=True)
 
-        logger.info(f"Loading fire model...")
+        logger.info("Loading fire model...")
 
         # Load net
         state = torch.load(model_path)

hloc/extractors/fire_local.py

@@ -1,11 +1,12 @@
-from pathlib import Path
 import subprocess
 import sys
+from pathlib import Path
+
 import torch
 import torchvision.transforms as tvf
 
-from ..utils.base_model import BaseModel
 from .. import logger
+from ..utils.base_model import BaseModel
 
 fire_path = Path(__file__).parent / "../../third_party/fire"
 
@@ -13,10 +14,6 @@ sys.path.append(str(fire_path))
 
 
 import fire_network
-from lib.how.how.stages.evaluate import eval_asmk_fire, load_dataset_fire
-
-from lib.asmk import asmk
-from asmk import io_helpers, asmk_method, kernel as kern_pkg
 
 EPS = 1e-6
 
@@ -44,18 +41,18 @@ class FIRe(BaseModel):
 
         # Config paths
         model_path = fire_path / "model" / conf["model_name"]
-        config_path = fire_path / conf["config_name"]
-        asmk_bin_path = fire_path / "model" / conf["asmk_name"]
+        config_path = fire_path / conf["config_name"]  # noqa: F841
+        asmk_bin_path = fire_path / "model" / conf["asmk_name"]  # noqa: F841
 
         # Download the model.
         if not model_path.exists():
             model_path.parent.mkdir(exist_ok=True)
             link = self.fire_models[conf["model_name"]]
-            cmd = ["wget", link, "-O", str(model_path)]
+            cmd = ["wget", "--quiet", link, "-O", str(model_path)]
             logger.info(f"Downloading the FIRe model with `{cmd}`.")
             subprocess.run(cmd, check=True)
 
-        logger.info(f"Loading fire model...")
+        logger.info("Loading fire model...")
 
         # Load net
         state = torch.load(model_path)

hloc/extractors/lanet.py

@@ -1,14 +1,17 @@
+import subprocess
 import sys
 from pathlib import Path
-import subprocess
+
 import torch
+from hloc import logger
 
 from ..utils.base_model import BaseModel
-from hloc import logger
+
+lib_path = Path(__file__).parent / "../../third_party"
+sys.path.append(str(lib_path))
+from lanet.network_v0.model import PointModel
 
 lanet_path = Path(__file__).parent / "../../third_party/lanet"
-sys.path.append(str(lanet_path))
-from network_v0.model import PointModel
 
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
@@ -26,11 +29,11 @@ class LANet(BaseModel):
             lanet_path / "checkpoints" / f'PointModel_{conf["model_name"]}.pth'
         )
         if not model_path.exists():
-            print(f"No model found at {model_path}")
+            logger.warning(f"No model found at {model_path}, start downloading")
         self.net = PointModel(is_test=True)
         state_dict = torch.load(model_path, map_location="cpu")
         self.net.load_state_dict(state_dict["model_state"])
-        logger.info(f"Load LANet model done.")
+        logger.info("Load LANet model done.")
 
     def _forward(self, data):
         image = data["image"]

hloc/extractors/netvlad.py

@@ -1,5 +1,6 @@
-from pathlib import Path
 import subprocess
+from pathlib import Path
+
 import numpy as np
 import torch
 import torch.nn as nn
@@ -7,8 +8,8 @@ import torch.nn.functional as F
 import torchvision.models as models
 from scipy.io import loadmat
 
-from ..utils.base_model import BaseModel
 from .. import logger
+from ..utils.base_model import BaseModel
 
 EPS = 1e-6
 
@@ -60,7 +61,7 @@ class NetVLAD(BaseModel):
         if not checkpoint.exists():
             checkpoint.parent.mkdir(exist_ok=True, parents=True)
             link = self.dir_models[conf["model_name"]]
-            cmd = ["wget", link, "-O", str(checkpoint)]
+            cmd = ["wget", "--quiet", link, "-O", str(checkpoint)]
             logger.info(f"Downloading the NetVLAD model with `{cmd}`.")
             subprocess.run(cmd, check=True)
 

hloc/extractors/r2d2.py

@@ -1,14 +1,15 @@
 import sys
 from pathlib import Path
+
 import torchvision.transforms as tvf
 
-from ..utils.base_model import BaseModel
 from hloc import logger
 
-base_path = Path(__file__).parent / "../../third_party"
-sys.path.append(str(base_path))
+from ..utils.base_model import BaseModel
+
 r2d2_path = Path(__file__).parent / "../../third_party/r2d2"
-from r2d2.extract import load_network, NonMaxSuppression, extract_multiscale
+sys.path.append(str(r2d2_path))
+from extract import NonMaxSuppression, extract_multiscale, load_network
 
 
 class R2D2(BaseModel):
@@ -35,7 +36,7 @@ class R2D2(BaseModel):
             rel_thr=conf["reliability_threshold"],
             rep_thr=conf["repetability_threshold"],
         )
-        logger.info(f"Load R2D2 model done.")
+        logger.info("Load R2D2 model done.")
 
     def _forward(self, data):
         img = data["image"]

hloc/extractors/rekd.py

@@ -1,11 +1,12 @@
 import sys
 from pathlib import Path
-import subprocess
+
 import torch
 
-from ..utils.base_model import BaseModel
 from hloc import logger
 
+from ..utils.base_model import BaseModel
+
 rekd_path = Path(__file__).parent / "../../third_party"
 sys.path.append(str(rekd_path))
 from REKD.training.model.REKD import REKD as REKD_
@@ -29,7 +30,7 @@ class REKD(BaseModel):
         self.net = REKD_(is_test=True)
         state_dict = torch.load(model_path, map_location="cpu")
         self.net.load_state_dict(state_dict["model_state"])
-        logger.info(f"Load REKD model done.")
+        logger.info("Load REKD model done.")
 
     def _forward(self, data):
         image = data["image"]

hloc/extractors/rord.py

@@ -3,9 +3,10 @@ from pathlib import Path
 import subprocess
 import torch
 
-from ..utils.base_model import BaseModel
 from hloc import logger
 
+from ..utils.base_model import BaseModel
+
 rord_path = Path(__file__).parent / "../../third_party"
 sys.path.append(str(rord_path))
 from RoRD.lib.model_test import D2Net as _RoRD
@@ -42,11 +43,10 @@ class RoRD(BaseModel):
                     subprocess.run(cmd, check=True)
                 except subprocess.CalledProcessError as e:
                     logger.error(f"Failed to download the RoRD model.")
-                    raise e
         self.net = _RoRD(
             model_file=model_path, use_relu=conf["use_relu"], use_cuda=False
         )
-        logger.info(f"Load RoRD model done.")
+        logger.info("Load RoRD model done.")
 
     def _forward(self, data):
         image = data["image"]

hloc/extractors/sfd2.py

@@ -0,0 +1,43 @@
+# -*- coding: UTF-8 -*-
+import sys
+from pathlib import Path
+
+import torchvision.transforms as tvf
+
+from .. import logger
+from ..utils.base_model import BaseModel
+
+pram_path = Path(__file__).parent / "../../third_party/pram"
+sys.path.append(str(pram_path))
+
+from nets.sfd2 import load_sfd2
+
+
+class SFD2(BaseModel):
+    default_conf = {
+        "max_keypoints": 4096,
+        "model_name": "sfd2_20230511_210205_resnet4x.79.pth",
+        "conf_th": 0.001,
+    }
+    required_inputs = ["image"]
+
+    def _init(self, conf):
+        self.conf = {**self.default_conf, **conf}
+        self.norm_rgb = tvf.Normalize(
+            mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]
+        )
+        model_fn = pram_path / "weights" / self.conf["model_name"]
+        self.net = load_sfd2(weight_path=model_fn).eval()
+
+        logger.info("Load SFD2 model done.")
+
+    def _forward(self, data):
+        pred = self.net.extract_local_global(
+            data={"image": self.norm_rgb(data["image"])}, config=self.conf
+        )
+        out = {
+            "keypoints": pred["keypoints"][0][None],
+            "scores": pred["scores"][0][None],
+            "descriptors": pred["descriptors"][0][None],
+        }
+        return out

hloc/extractors/sift.py

@@ -4,14 +4,15 @@ import cv2
 import numpy as np
 import torch
 from kornia.color import rgb_to_grayscale
-from packaging import version
 from omegaconf import OmegaConf
+from packaging import version
 
 try:
     import pycolmap
 except ImportError:
     pycolmap = None
 from hloc import logger
+
 from ..utils.base_model import BaseModel
 
 
@@ -140,7 +141,7 @@ class SIFT(BaseModel):
                 f"Unknown backend: {backend} not in "
                 f"{{{','.join(backends)}}}."
             )
-        logger.info(f"Load SIFT model done.")
+        logger.info("Load SIFT model done.")
 
     def extract_single_image(self, image: torch.Tensor):
         image_np = image.cpu().numpy().squeeze(0)

hloc/extractors/superpoint.py

@@ -1,10 +1,12 @@
 import sys
 from pathlib import Path
+
 import torch
 
-from ..utils.base_model import BaseModel
 from hloc import logger
 
+from ..utils.base_model import BaseModel
+
 sys.path.append(str(Path(__file__).parent / "../../third_party"))
 from SuperGluePretrainedNetwork.models import superpoint  # noqa E402
 
@@ -43,7 +45,7 @@ class SuperPoint(BaseModel):
         if conf["fix_sampling"]:
             superpoint.sample_descriptors = sample_descriptors_fix_sampling
         self.net = superpoint.SuperPoint(conf)
-        logger.info(f"Load SuperPoint model done.")
+        logger.info("Load SuperPoint model done.")
 
     def _forward(self, data):
         return self.net(data, self.conf)

hloc/extractors/xfeat.py

@@ -1,6 +1,7 @@
 import torch
-from pathlib import Path
+
 from hloc import logger
+
 from ..utils.base_model import BaseModel
 
 
@@ -18,7 +19,7 @@ class XFeat(BaseModel):
             pretrained=True,
             top_k=self.conf["max_keypoints"],
         )
-        logger.info(f"Load XFeat(sparse) model done.")
+        logger.info("Load XFeat(sparse) model done.")
 
     def _forward(self, data):
         pred = self.net.detectAndCompute(

hloc/match_dense.py

@@ -1,9 +1,11 @@
+from types import SimpleNamespace
+
+import cv2
 import numpy as np
 import torch
 import torchvision.transforms.functional as F
-from types import SimpleNamespace
+
 from .extract_features import read_image, resize_image
-import cv2
 
 device = "cuda" if torch.cuda.is_available() else "cpu"
 

hloc/match_features.py

@@ -1,18 +1,19 @@
 import argparse
-from typing import Union, Optional, Dict, List, Tuple
-from pathlib import Path
 import pprint
+from functools import partial
+from pathlib import Path
 from queue import Queue
 from threading import Thread
-from functools import partial
-from tqdm import tqdm
+from typing import Dict, List, Optional, Tuple, Union
+
 import h5py
+import numpy as np
 import torch
+from tqdm import tqdm
 
-from . import matchers, logger
+from . import logger, matchers
 from .utils.base_model import dynamic_load
 from .utils.parsers import names_to_pair, names_to_pair_old, parse_retrieval
-import numpy as np
 
 """
 A set of standard configurations that can be directly selected from the command
@@ -162,6 +163,13 @@ confs = {
             "match_threshold": 0.2,
         },
     },
+    "imp": {
+        "output": "matches-imp",
+        "model": {
+            "name": "imp",
+            "match_threshold": 0.2,
+        },
+    },
 }
 
 

hloc/matchers/adalam.py

@@ -1,10 +1,9 @@
 import torch
-
-from ..utils.base_model import BaseModel
-
 from kornia.feature.adalam import AdalamFilter
 from kornia.utils.helpers import get_cuda_device_if_available
 
+from ..utils.base_model import BaseModel
+
 
 class AdaLAM(BaseModel):
     # See https://kornia.readthedocs.io/en/latest/_modules/kornia/feature/adalam/adalam.html.

hloc/matchers/aspanformer.py

@@ -1,17 +1,16 @@
+import subprocess
 import sys
-import torch
-from ..utils.base_model import BaseModel
-from ..utils import do_system
 from pathlib import Path
-import subprocess
+
+import torch
 
 from .. import logger
+from ..utils.base_model import BaseModel
 
 sys.path.append(str(Path(__file__).parent / "../../third_party"))
 from ASpanFormer.src.ASpanFormer.aspanformer import ASpanFormer as _ASpanFormer
 from ASpanFormer.src.config.default import get_cfg_defaults
 from ASpanFormer.src.utils.misc import lower_config
-from ASpanFormer.demo import demo_utils
 
 aspanformer_path = Path(__file__).parent / "../../third_party/ASpanFormer"
 
@@ -85,7 +84,7 @@ class ASpanFormer(BaseModel):
             "state_dict"
         ]
         self.net.load_state_dict(state_dict, strict=False)
-        logger.info(f"Loaded Aspanformer model")
+        logger.info("Loaded Aspanformer model")
 
     def _forward(self, data):
         data_ = {

hloc/matchers/cotr.py

@@ -1,19 +1,19 @@
-import sys
 import argparse
-import torch
-import warnings
-import numpy as np
+import sys
 from pathlib import Path
+
+import numpy as np
+import torch
 from torchvision.transforms import ToPILImage
+
 from ..utils.base_model import BaseModel
 
 sys.path.append(str(Path(__file__).parent / "../../third_party/COTR"))
-from COTR.utils import utils as utils_cotr
-from COTR.models import build_model
-from COTR.options.options import *
-from COTR.options.options_utils import *
-from COTR.inference.inference_helper import triangulate_corr
 from COTR.inference.sparse_engine import SparseEngine
+from COTR.models import build_model
+from COTR.options.options import *  # noqa: F403
+from COTR.options.options_utils import *  # noqa: F403
+from COTR.utils import utils as utils_cotr
 
 utils_cotr.fix_randomness(0)
 torch.set_grad_enabled(False)
@@ -33,7 +33,7 @@ class COTR(BaseModel):
 
     def _init(self, conf):
         parser = argparse.ArgumentParser()
-        set_COTR_arguments(parser)
+        set_COTR_arguments(parser)  # noqa: F405
         opt = parser.parse_args()
         opt.command = " ".join(sys.argv)
         opt.load_weights_path = str(

hloc/matchers/dkm.py

@@ -1,10 +1,12 @@
+import subprocess
 import sys
 from pathlib import Path
+
 import torch
 from PIL import Image
-import subprocess
-from ..utils.base_model import BaseModel
+
 from .. import logger
+from ..utils.base_model import BaseModel
 
 sys.path.append(str(Path(__file__).parent / "../../third_party"))
 from DKM.dkm import DKMv3_outdoor
@@ -37,11 +39,11 @@ class DKMv3(BaseModel):
         if not model_path.exists():
             model_path.parent.mkdir(exist_ok=True)
             link = self.dkm_models[conf["model_name"]]
-            cmd = ["wget", link, "-O", str(model_path)]
+            cmd = ["wget", "--quiet", link, "-O", str(model_path)]
             logger.info(f"Downloading the DKMv3 model with `{cmd}`.")
             subprocess.run(cmd, check=True)
         self.net = DKMv3_outdoor(path_to_weights=str(model_path), device=device)
-        logger.info(f"Loading DKMv3 model done")
+        logger.info("Loading DKMv3 model done")
 
     def _forward(self, data):
         img0 = data["image0"].cpu().numpy().squeeze() * 255