| Milestone | Date | Status |
|---|---|---|
| Preliminary submission deadline | 2026-09-01 23:59:59 | DEADLINE |
| Announcement of finalists | 2026-09-03 | GATE |
| Finalist submission deadline | 2026-09-15 23:59:59 | CONDITIONAL DEADLINE |
Strong model controls
Co-DINO1280 and RT-DETRv4-X1280 are accepted. Co-DINO1280 demonstrated a local and hidden scale-transfer gain.
MI-DETR Effective-Scale AUG V2
Finish the 12-epoch AWS L4 run before making a causal replacement decision.
Team system handoff
Reconstruct the approximately 0.7701 hidden pipeline before treating it as fully canonical local authority.
Current decision tree · Preliminary phase
1 · Finish MI-DETR1280 Effective-Scale AUG V2Record the complete 12-epoch trajectory and compare against MI-DETR1024 = 0.7154639
2 · Complete teammate artifact handoffReproduce the approximately 0.7701 team system with exact artifacts, preprocessing, scoring, and submission provenance
3 · Retire Co-DINO Full-Aug V1Preserve failure evidence; do not continue the scale-cancelling composed V1
4 · Run Co-DINO Effective-Scale AUG V2Fresh Objects365 initialization; no final scale-normalizing resize; correct partial accumulation window
5 · If MI-DETR V2 underperformsRun full MI-DETR1280 scale-control; do not auto-start a 24-epoch V2 retrain
6 · Freeze strongest Co-DINO + MI-DETR variantsChoose by canonical validation, hidden evidence, and contribution to the reproduced final system
7 · Test replacements inside rescorer pipelineMeasured substitutions only; retain RT-DETRv4 if it stays additive
8 · Internal freezeTarget: 31 August
9 · Final preliminary submission · DEADLINE2026-09-01 23:59:59 Taiwan time
| Variant / branch | Current state | Next sufficient action |
|---|---|---|
| MI-DETR1280 Effective-Scale AUG V2 | ACTIVE · best 0.708067 at epoch 6 | finish epochs 10–12 |
| Co-DINO1280 Full-Aug V1 | FAILED · SUPERSEDED | retire; do not continue |
| Co-DINO1280 Effective-Scale AUG V2 | PLANNED / NEXT | fresh controlled run after bounded GPU qualification |
| MI-DETR1280 scale-control | INCOMPLETE / DEPRIORITIZED | run full control only if V2 remains clearly below 1024 |
| RT-DETRv4-X1280 | COMPLETE · ACCEPTED | retain only if additive in rescorer pipeline |
| Broad detector sweep | NOT CURRENT PRIORITY | avoid blind additions |
strong model completion → ensemble optimization → crop rescoring → final submission system
| Item | Current authority |
|---|---|
| Component | ConvNeXt-Tiny crop rescorer · PROVEN HIGH-VALUE COMPETITION COMPONENT |
| Observed evidence | approximately 0.7701 AIdea hidden · team pipeline · reproduction handoff in progress |
| Fully controlled fallback | Co-DINO1280 + MI-DETR1024 WBF · local 0.7468375071854423 · hidden 0.7587924 |
| Handoff contents | checkpoints · validation/preliminary predictions · exact ensemble recipe · crop preprocessing · score formula · final submission SHA · documentation |
| Artifact convention | small provenance in Git; large prediction/checkpoint artifacts in LFS/shared storage; datasets externally |
IF SELECTED AS FINALIST · starts after 3 September
Finalist announcement · GATE2026-09-03 · this entire branch happens only if selected
Stronger crop-rescorer refinementRevalidate against the frozen finalist detector and fusion distribution
Leakage-safe calibration / simple stackingOnly if a safe OOF or frozen fit/selection strategy exists and measured gain justifies complexity
Measured class/area gating or tiny specialistRetain only if evidence supports a net final-system benefit
Additional controlled Effective-Scale experimentsOnly if a specific unresolved hypothesis remains
Cross-team ensemble ablationCompare shared artifacts; do not assume components transfer without measurement
Final submission · CONDITIONAL DEADLINE2026-09-15 23:59:59 Taiwan time
| Decision | Authority |
|---|---|
| Conceptual input | Co-DINO + MI-DETR + optional RT-DETRv4 → WBF → crop rescorer → Meta Stacker |
| Possible features | class ID · area fraction · aspect ratio · detector confidences · WBF score · supporter count · IoU/agreement · crop score · optional density |
| Output | confidence / ranking only · no class reassignment · no box geometry modification |
| Initial complexity | logistic regression or small gradient-boosted tree; shallow MLP only if justified |
| Leakage-safe fitting | OOF detector predictions if practical, or a carefully frozen fit/selection partition |
| Stop rule | skip if safe fitting is not cheap; retain only for material frozen-validation gain over WBF + crop rescorer |