[CVPR 2025] MINIMA: Modality Invariant Image Matching
active 2024-12-25 → 2026-04-18 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 480 days from 2024-12-25 to 2026-04-18. Pushes: 47 total, peak 19 in a day. Pull requests: 0 total, peak 0 in a day. Issues: 50 total, peak 4 in a day. Comments: 73 total, peak 8 in a day. Stars: 471 total, peak 28 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| LSXI7 | 83 | 44 | 0 | 39 |
| StaRainJ | 8 | 3 | 0 | 5 |
| skbhat | 6 | 0 | 0 | 2 |
| kevinhelvig | 6 | 0 | 0 | 4 |
| zj12345648 | 5 | 0 | 0 | 2 |
| Sharpiless | 5 | 0 | 0 | 1 |
| dahexiangdong6 | 4 | 0 | 0 | 1 |
| Shexiaox | 4 | 0 | 0 | 3 |
| maikousmisi | 4 | 0 | 0 | 3 |
| HarborC | 3 | 0 | 0 | 2 |
| ZhonghuaYi | 3 | 0 | 0 | 2 |
| annieiscc | 2 | 0 | 0 | 0 |
| chrisgank | 2 | 0 | 0 | 1 |
| donkeymouse | 2 | 0 | 0 | 0 |
| leoluxxx | 2 | 0 | 0 | 0 |
| JINWEI8729 | 2 | 0 | 0 | 1 |
| abox2023 | 2 | 0 | 0 | 1 |
| Osirising | 2 | 0 | 0 | 1 |
| miaoqingran | 1 | 0 | 0 | 0 |
| NielsRogge | 1 | 0 | 0 | 0 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#51Sj-Yuan2026-01-22 13:24Can't import Swift because of a problem that 'type' object is not subscriptable
- Issue#57cuffsword2026-01-06 14:18Training with your own data
- Issue comment#56LSXI72025-12-16 06:30Alternative feature extractors other than Superpoint
- Issue#53O1dCoo12025-11-04 05:35Request for DIODE Dataset with .png Depth Images (.npy incompatible with test script)
- Issue comment#53LSXI72025-11-01 08:18Request for DIODE Dataset with .png Depth Images (.npy incompatible with test script)
- Issue comment#7DruncBread2025-10-18 03:21请问，lightglue在微调时有训练superpoint吗？
- Issue comment#49LSXI72025-10-08 05:59Problem in generating infrared modality by using data-engine
- Issue comment#48LSXI72025-10-08 05:49About GPU memory during training
- Issue#49fallenskies002025-09-30 09:29Problem in generating infrared modality by using data-engine
- Issue comment#44LSXI72025-09-29 06:29A minor issue with a file path location
- Issue comment#44BuWen-TP2025-09-13 14:31A minor issue with a file path location
- Issue comment#45Shexiaox2025-08-15 14:17Performance issues with local demonstrations
- Issue comment#29Shexiaox2025-08-15 09:34Hugging Face demo performs better than demo.py
- Issue comment#9JINWEI87292025-08-15 09:20Inconsistent Matching Results Across Different Runs and Between Local/Online Environments
- Issue comment#9Shexiaox2025-08-15 04:29Inconsistent Matching Results Across Different Runs and Between Local/Online Environments
- Issue#44Shexiaox2025-08-15 03:50A minor issue with a file path location
- Issue#42sbucaille2025-07-31 20:16MINIMA is supported by HuggingFace Transformers library
- Issue comment#41wangtan23672025-07-23 15:52There is no “src.utils.load_model” file in demo.py
- Issue#31annieiscc2025-06-26 08:46数据集问题求教
- Issue#35leoluxxx2025-06-25 01:59The demo in huggingface performs much better than the demo in the source code
- Issue comment#36LSXI72025-06-23 07:21The questions of the Event-Image test set.
- Issue#37CrazyCoder762025-06-21 13:21onnx export issue
- Issue comment#36ZhonghuaYi2025-06-21 05:08The questions of the Event-Image test set.
- Issue comment#35LSXI72025-06-20 06:44The demo in huggingface performs much better than the demo in the source code
- Issue#35leoluxxx2025-06-13 01:51The demo in huggingface performs much better than the demo in the source code
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 471 stars here means stars gained during the window, not the repo's star count.