XFeat
XFeat (Verlab, CVPR 2024): fast sparse keypoints and descriptors, matched by mutual nearest neighbours or LighterGlue; output as keypoints plus index pairs (the form COLMAP takes). Apache-2.0 code and weights.
experimental Fork of verlab/accelerated_features@e92685f · backends: portable, kernels
- Paper XFeat: Accelerated Features for Lightweight Image Matching CVPR 2024 · first released 30 Apr 2024
Paper 2024 Added in 0.2.0 Evaluated with 0.2.0
Use it
pip install fastforks Lean install with only what this model needs: pip install fastforks-core
from fastforks import matching
model = matching.load("xfeat") # kernels on a supported GPU, else portable PyTorch
result = model(image_a, image_b)
result.keypoints1 # also: keypoints2, pairs, scores
result.backend # what ran: mode, device, dtype, kernels
points1, points2 = result.matches(5000) # matched pixels, sampled where the model is certain
pose = matching.estimate_pose(points1, points2, K1, K2) # relative pose on the GPU (RANSAC); K = intrinsics [3, 3]
pose.R, pose.t, pose.inliers # rotation, unit translation, inlier mask fastforks bench xfeat Times the model on this machine. load(…, mode="portable") or mode="kernels" picks a backend; usage guide.
Speed and memory
Every backend on one GPU (RTX 5060 Ti 16 GB) and the same input, the authors’ code included. Kernels runs fp32. Portable runs fp32 on CUDA.
Input: synthetic, 768x1024 (the task’s standard size)
| Backend | dtype | Time per call (ms) | vs original fp16 | Peak memory allocated / reserved | Load (s) | First call (s) |
|---|---|---|---|---|---|---|
| Original fp32 | fp32 | not measured | – | – | – | – |
| Original fp16torch.autocast(float16) around the call; upstream XFeat.detectAndCompute + mnn, top_k 4096 | fp16 | 8.89 | – | 128 MiB/174 MiB | 0.3 | 0.32 |
| Portable fp32 | fp32 | not measured | – | – | – | – |
| Portable | fp32 | 7.17 | 1.2× | 122 MiB/154 MiB | 0.2 | 0.74 |
| Kernels | not measured | – | – | – | – | |
| Portable + torch.compile | 5.65 | 1.6× | 186 MiB/212 MiB | 31 (build) | – | |
| TensorRTplanned for 0.2.1 | fails | – | – | – | – |
Measured on RTX 5060 Ti 16 GB, torch 2.14.1+cu130, 2026-11-12 to 2026-11-13. First call, 3 warm-up, peak reset, median of 3 rounds x median of 5. Ratios divide two
times of this table (same GPU, same input). Original fp16 is the authors’ own precision option, else torch.autocast. torch.compile and TensorRT: build time instead of load time. TensorRT memory (engine, not comparable) is the GPU memory the engine took, read from the driver because TensorRT allocates outside PyTorch, so it can't be compared with the PyTorch peaks of the other rows.
Other devices
| Device | Backend | dtype | ms / call | Peak memory | Output vs reference | Input |
|---|---|---|---|---|---|---|
| Apple M1 Pro 16 GB MPS, torch 2.14.1, 2026-11-13 | Original | fp32 | 45.2 | 1,549 MiB process | – | 768x1024 |
| Portable | fp32 | 41.7 | 1,471 MiB process | matches | 768x1024 | |
| Kernels | fp32 | not implemented | – | 768x1024 | ||
CUDA memory is torch’s peak allocated / reserved; Apple has no such counter, so the Mac rows show the process footprint (weights, Python and torch included). “Output vs reference” compares the device’s output with the stored golden output under the card’s tolerance; – = not checked on that device. Kernels “not implemented” = no kernels for this model on that backend yet: it runs portable there.
Licence and weights
Commercial use allowed
Our reading of the licences, not legal advice.
Weights
xfeat.pt 6.2 MB· hosted on GitHubfastforks downloads and checks these files on first use (fastforks download xfeat).
Accuracy
The test set through the authors’ code and through ours: same checkpoint, same protocol. The distance between the original’s fp32 and fp16 runs is what half precision alone costs; our runs should stay within it.
MegaDepth-1500
relative pose AUC@5 / @10 / @20 degrees
| Run | dtype | AUC@5° ↑ | AUC@10° ↑ | AUC@20° ↑ | pairs | ms / pair |
|---|---|---|---|---|---|---|
| Original fp32 | not measured | |||||
| Original fp16 | not measured | |||||
| Portable fp32 | fp32 | 26.2 | 42.2 | 57.5 | 1500 | 15.5 |
| Portable | fp32 | 26.2 | 42.2 | 57.5 | 1500 | 15.5 |
| Kernels | not measured | |||||
The original code was not run on this dataset, so the table cannot show the distance to it.
Dense matchers: 5000 matches per pair sampled from the warp by certainty (upstream's balanced KDE sampling); sparse matchers: all their matches of one call (LightGlue, XFeat: LightGlue's MegaDepth-1500 setting; LoFTR, DeDoDe: their card defaults). OpenCV essential-matrix RANSAC (0.5 px threshold, normalised by the mean focal length); AUC of the larger of the rotation and translation-direction errors. Measured on RTX 5060 Ti, torch 2.14.1+cu130, fastforks ec7b1307+por, 2026-11-12.
MegaDepth-1500
every 5th pair, relative pose AUC@5 / @10 / @20 degrees
| Run | dtype | AUC@5° ↑ | AUC@10° ↑ | AUC@20° ↑ | Δ AUC@5° vs fp32 original | pairs | ms / pair |
|---|---|---|---|---|---|---|---|
| Original fp32 | fp32 | 29.1 | 46.1 | 60.8 | reference | 300 | 15.4 |
| Original fp16 | fp16 | 28.4 | 44.6 | 60.1 | −0.8 | 300 | 12.6 |
| Portable fp32 | fp32 | 29.1 | 46.1 | 60.8 | ±0.0 | 300 | 15.7 |
| Portable | fp32 | 29.1 | 46.1 | 60.8 | ±0.0 | 300 | 15.7 |
| Kernels | not measured | ||||||
Dense matchers: 5000 matches per pair sampled from the warp by certainty (upstream's balanced KDE sampling); sparse matchers: all their matches of one call (LightGlue, XFeat: LightGlue's MegaDepth-1500 setting; LoFTR, DeDoDe: their card defaults). OpenCV essential-matrix RANSAC (0.5 px threshold, normalised by the mean focal length); AUC of the larger of the rotation and translation-direction errors. The same 300 pairs for all modes, so float32 and bfloat16 compare pair for pair. Original = the authors’ code, same checkpoint and protocol. Measured on original fp32, original fp16: RTX 5060 Ti, torch 2.14.1+cu130, 2026-11-12; portable fp32, portable: RTX 5060 Ti, torch 2.14.1+cu130, fastforks ec7b1307+por, 2026-11-12.
Relative pose on the GPU
A dense matcher is usually followed by RANSAC on its matches to get the camera motion. fastforks does that step on the
GPU too: result.matches(n) samples matched points where the model is certain, and matching.estimate_pose runs essential-matrix RANSAC on them, many image pairs per call, with no copy to
the CPU. Like the models it has a portable (plain PyTorch) and a kernels (CUDA) backend.
| Solver | Runs on | Threshold (px) | Pose AUC@5 / @10 / @20 | Pairs per second |
|---|---|---|---|---|
| fastforks estimate_pose, kernels | GPU, batches of 500 pairs | 1.0 | 69.9 / 81.5 / 89.3 | 18,446 |
| fastforks estimate_pose, portable | GPU, batches of 500 pairs | 1.0 | 69.6 / 81.2 / 89.1 | 418 |
| PoseLib 2.0.5 | CPU, 16 processes | 1.0 | 70.0 / 81.6 / 89.5 | 325 |
| OpenCV USAC (MAGSAC++) | CPU, 16 processes | 0.5 | 65.7 / 78.6 / 87.3 | 271 |
| OpenCV RANSAC | CPU, 16 processes | 0.5 | 62.9 / 76.9 / 86.4 | 323 |
MegaDepth-1500, RoMa v2 matches (5000 per pair), 5 samplings, the same matches for every solver, on one machine: RTX 5060 Ti and a Ryzen 7 5800X (16 threads). Each solver runs at the inlier threshold where it is most accurate. Pairs per second is wall-clock throughput over all 7500 poses: the CPU solvers run one pair per call in 16 single-threaded processes, the GPU solver takes 500 pairs per call. One pair alone takes about 1.7 ms with the kernels, 29 ms with PoseLib and 24 ms with OpenCV RANSAC on one core. The accuracy table above stays on OpenCV RANSAC, as in the papers, so it is comparable with published numbers; this table is the only place the GPU solver is used.
Closeness to the original
Beyond the test-set scores: the raw output of ours against the original’s fp32 output on the same input, pixel by pixel. CPU, float32: max |diff| <= 3e-05 vs the original. GPU, float32 (the default), portable and kernels: mean |diff| <= 0.0001 and p99 <= 0.001 (absolute error of the output keypoints1).
| Check | Tolerance | Measured |
|---|---|---|
| CPU fp32, max |diff| | ≤ 0.00003 | not recorded |
| cpu_min_recall | ≤ 0.99 | not recorded |
| cpu_min_match | ≤ 0.99 | not recorded |
| GPU default dtypes, mean |diff| | ≤ 0.0001 | 0 |
| GPU default dtypes, p99 |diff| | ≤ 0.001 | 0 |
| gpu_min_recall | ≤ 0.999 | not recorded |
| gpu_min_match | ≤ 0.993 | not recorded |
Recorded for xfeat on RTX 5060 Ti, torch 2.10.0+cu130, 2026-11-17, with these kernels: keypoint_nms
Details
XFeat (Potje et al., CVPR 2024): lightweight sparse feature matching. A small CNN finds up to 4096 keypoints per image with 64-dimensional descriptors; they are matched as mutual nearest neighbours or with LighterGlue, a smaller LightGlue trained for XFeat. Returns both keypoint sets, the index pairs of the matches and their confidence.
- Input / output
- Input: two RGB images of one scene. Output: for every pixel of the first image its position in the second (x, y in pixels) and a certainty that the pixel is visible in both.
What the numbers mean, and which models need the camera focal length
Sparse matchers (LightGlue) return the keypoints of both images and the index pairs of the matched ones instead. To reconstruct camera poses from an image folder with these matchers, see [structure from motion](../sfm.md).
- Input size
- any H x W >= 32, padded to a multiple of 32; tested up to 3000x4000 (1,047 MiB allocated)
- Options
max_keypoints=4096>= 1· changes the outputlong_side=0>= 0· changes the output
- Outputs
keypoints1,keypoints2,pairs,scores- Checkpoint
matcher="mnn",detection_threshold=0.05- CUDA kernels
keypoint_nms
- Apple (MPS)
- fp32
Limitations
- Sparse output: no dense warp. `result.matches()` gives the matched points for `estimate_pose`; `pairs` are the index pairs pycolmap / COLMAP databases take.
- One image pair per call internally (batches run pair by pair); the keypoint counts differ per image.
- No exportable module: the keypoint counts are data-dependent (`torch.compile` runs with graph breaks; ONNX / TensorRT export fails).
- `xfeat` (mutual nearest neighbours) has no learned confidence: its scores are the descriptors' cosine similarities.
Full reference, generated from the repository
XFeat (Verlab, CVPR 2024): fast sparse keypoints and descriptors, matched by mutual nearest neighbours or LighterGlue; output as keypoints plus index pairs (the form COLMAP takes). Apache-2.0 code and weights.
Commercial use: yes.
| variant | input | backend | time (ms) | peak memory (MiB, allocated / reserved) |
|---|---|---|---|---|
xfeat | synthetic (768x1024) | original portable kernels portable + torch.compile TensorRT fp16 | 8.89 7.17 – 5.65 planned for 0.2.1 | 128 / 174 122 / 154 – 186 / 212 – |
xfeat-lighterglue | synthetic (768x1024) | original portable kernels portable + torch.compile TensorRT fp16 | 23.6 21.6 – 16.2 planned for 0.2.1 | 448 / 478 445 / 476 – 262 / 278 – |
Details
- Original code: https://github.com/verlab/accelerated_features at
e92685f57f83 - Code licence: Apache-2.0 (permissive); files:
LICENSE - Weights licence: Apache-2.0
- Licence evidence: https://github.com/verlab/accelerated_features/blob/e92685f57f8318b18725c5c8c0bd28c7fe188d9a/LICENSE ; https://github.com/verlab/accelerated_features/blob/e92685f57f8318b18725c5c8c0bd28c7fe188d9a/README.md#license
- Training data: MegaDepth (Internet photo collections with SfM / MVS depth), COCO 2017 (train, instance annotations) (XFeat: MegaDepth and synthetic warps of a 20k-image COCO 2017 subset (README training section). LighterGlue: trained with GlueFactory (README; MegaDepth, data not otherwise stated).)
- Options:
max_keypoints= 4096 (>= 1),long_side= 0 (>= 0) - Kernel libraries:
keypoint_nms - Variants:
xfeat,xfeat-lighterglue - Not in this release (planned for 0.2.1):
xfeatONNX / TensorRT export: torch.export stops at a data-dependent branch (NMS / keypoint selection);xfeat-lighterglueONNX / TensorRT export: torch.export stops at a data-dependent branch (NMS / keypoint selection) - Outputs: keypoints1, keypoints2, pairs, scores
- Input size: any H x W >= 32, padded to a multiple of 32
- Tested up to (portable, fp16, RTX 5060 Ti 16 GB):
xfeat3000x4000 (1.0 GiB peak allocated, 1.2 reserved);xfeat-lighterglue3000x4000 (1.0 GiB peak allocated, 1.2 reserved) xfeat768x1024, first call: portable + torch.compile 31 s (cold Inductor cache).xfeat-lighterglue768x1024, first call: portable + torch.compile 40 s (cold Inductor cache).xfeatoriginal 768x1024: torch.autocast(float16) around the call; upstream XFeat.detectAndCompute + mnn, top_k 4096xfeat-lighterglueoriginal 768x1024: torch.autocast(float16) around the call; upstream XFeat.detectAndCompute + lighterglue, top_k 4096xfeatload / first call: original 1.51 s / 0.629 s, portable 0.16 s / 0.74 sxfeat-lighterglueload / first call: original 1.46 s / 0.692 s, portable 0.25 s / 0.681 s- Protocol, first call, 3 warm-up, peak reset, median of 3 rounds x median of 5
Measured
Dense matchers: 5000 matches per pair sampled from the warp by certainty (upstream's balanced KDE sampling); sparse matchers: all their matches of one call (LightGlue, XFeat: LightGlue's MegaDepth-1500 setting; LoFTR, DeDoDe: their card defaults). OpenCV essential-matrix RANSAC (0.5 px threshold, normalised by the mean focal length); AUC of the larger of the rotation and translation-direction errors. The same 300 pairs for all modes, so float32 and bfloat16 compare pair for pair. Original = the original code named under Details, same checkpoint and protocol.
| variant | dataset | original fp32 | original fp16 | portable | note |
|---|---|---|---|---|---|
xfeat | MegaDepth-1500, relative pose AUC@5 / @10 / @20 degrees | 26.2 / 42.2 / 57.5 | |||
xfeat | MegaDepth-1500, every 5th pair, relative pose AUC@5 / @10 / @20 degrees | 29.1 / 46.1 / 60.8 | 28.4 / 44.6 / 60.1 | 29.1 / 46.1 / 60.8 | |
xfeat-lighterglue | MegaDepth-1500, relative pose AUC@5 / @10 / @20 degrees | 39.7 / 57.5 / 72.8 | |||
xfeat-lighterglue | MegaDepth-1500, every 5th pair, relative pose AUC@5 / @10 / @20 degrees | 42.0 / 59.4 / 74.3 | 42.7 / 60.1 / 74.1 | 42.0 / 59.4 / 74.3 |
Measured on NVIDIA GeForce RTX 5060 Ti, torch 2.14.1+cu130, fastforks ec7b1307+port/p1-match, 2026-11-12.
Parity
CPU, float32: max |diff| <= 3e-05 vs the original. GPU, float32 (the default), portable and kernels: mean |diff| <= 0.0001 and p99 <= 0.001 (absolute error of the output keypoints1).
| variant | kernels error vs original, mean / p99 | GPU, torch | kernels used |
|---|---|---|---|
xfeat | 0.0 / 0.0 | RTX 5060 Ti, torch 2.10.0+cu130 | keypoint_nms@3cfcd62bc9a5:sm_120 |
xfeat-lighterglue | 0.0 / 0.0 | RTX 5060 Ti, torch 2.10.0+cu130 | keypoint_nms@3cfcd62bc9a5:sm_120 |
Limitations
- Sparse output: no dense warp.
result.matches()gives the matched points forestimate_pose;pairsare the index pairs pycolmap / COLMAP databases take. - One image pair per call internally (batches run pair by pair); the keypoint counts differ per image.
- No exportable module: the keypoint counts are data-dependent (
torch.compileruns with graph breaks; ONNX / TensorRT export fails). xfeat(mutual nearest neighbours) has no learned confidence: its scores are the descriptors' cosine similarities.
Example output
No example rendered for this checkpoint yet.