Pikaliov/caption-test
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Initial commit: caption quality test on UAV-VisLoc

Browse Source 
Self-contained experimental track validating generated text captions
via retrieval R@1 lift on UAV-VisLoc.

Architecture: GeoRSCLIP ViT-B/32 dual encoder, 512-dim shared space.
Loss: 4-term InfoNCE (img-img + sat-cap + drone-cap + cap-cap)
      with cosine temperature decay, PALW-like curriculum.
Metric: delta R@1 (with text - without text) >= +3% => PASS.

Gin-configured (balanced / baseline_no_text / text_heavy variants).
Follows NADEZHDA code style.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Pikaliov
2026-04-17 00:04:46 +03:00
commit 2ce4017ebd
18 changed files with 1864 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
conf/text_heavy.gin Normal file
View File

@@ -0,0 +1,11 @@
# Text-heavy configuration — stress test of caption contribution.
# L = 0.5 * L_img_img + 0.5 * L_sat_cap + 0.5 * L_drone_cap + 0.2 * L_cap_cap
include 'balanced.gin'
MultiTermInfoNCE.lambda_ii = 0.5
MultiTermInfoNCE.lambda_sc_max = 0.5
MultiTermInfoNCE.lambda_dc_max = 0.5
MultiTermInfoNCE.lambda_cc_max = 0.2
TrainConfig.output_dir = "out/caption_test/text_heavy"