5 Commits

Author SHA1 Message Date
pikaliov
95f41a4401 Fix gin config loading + source filter, docs updates
- config_loader: add base_gin_files() to exclude dataset-specific
  pipeline_*.gin variants from default `python -m src.main` load
  (pipeline_uav_visloc.gin was overriding pipeline.gin)
- main.py: use base_gin_files() in the --gin override branch too
- pipeline.gin: source None (both DB + query) — was 'query',
  which silently dropped all DB satellite crops from discovery
- README/CLAUDE/docs: sync source default, tensor format spec

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-07-11 17:27:51 +03:00
pikaliov
d1ddb40036 Update segmentation.gin to unified 17 classes
All datasets now use the same 17-class prompt list for transfer
learning compatibility. World-UAV segmentation needs regeneration.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-17 22:08:56 +03:00
pikaliov
6f04856b6b Add UAV_VisLoc annotation script + support drone folder naming
- Add scripts/run_uav_visloc.py for UAV_VisLoc_processed dataset (81K images)
- Add gin config for UAV_VisLoc (pipeline_uav_visloc.gin)
- Fix source filter and is_query_record to recognize "drone" folder
  (UAV_VisLoc uses "drone" instead of "query")

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-17 17:18:38 +03:00
pikaliov
f3cb18ac4d Add SafeTensors consolidation stage for zero-copy tensor loading
Bundle all per-image modalities (depth, edge, chm, segm) into a single
.safetensors file for fast training DataLoader reads (~0.1ms zero-copy mmap
vs ~5ms for 4x PNG). Adds consolidate stage after main pipeline stages,
save_safetensors/cleanup_npy config flags, resume support, and 10 new tests.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-17 16:53:29 +03:00
pikaliov
686db62c25 Initial commit: World-UAV annotation pipeline
4-modality annotation pipeline (depth, edges, segmentation, chmv2) for 973K
drone/satellite images. SegEarth-OV3 open-vocabulary segmentation with 11
classes optimized for cross-view geo-localization.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-16 11:22:01 +03:00