refactor: remove GPT/mission/relevance system, keep CV-only base

feat: expose step param on /detect/async, default to 7

feat(ttfs): instrument segmentation time-to-first-stream

refactor(gsam2): make SAM2 detector-agnostic

perf: increase production keyframe step from 20 to 60

perf: set SAM2 num_maskmem to 7 (trained default) in production

fix: replace gated GSAM2 streaming with adaptive-rate publisher thread

perf: set num_maskmem=3 for all inference paths

feat: add detailed timing metrics to GroundedSAM2 segmenter

feat: add job processing timing and latency measurements to backend and frontend

Fix GSAM2 GPT writer state safety and background call args

Remove SAM3 and standardize segmentation on Grounded-SAM2

Event-driven MJPEG streaming, GPT concurrency semaphore, pre-resize frames in publisher

Simplify first-frame processing, replace depth pre-scan with incremental stats, add GPT enrichment thread

Jobs: forward cached GPT results, reduce stream buffer to 60, wire up first_frame_gpt_results

feat: Add mission-driven object relevance abstractions

feat: Real-time radar directionality via track sync

feat: Continuous object tracking, speed estimation, and overlay syncing

Optimize streaming buffer and implement parallel depth pipeline

Fix syntax error in background job

Enable MJPEG streaming for live video feedback

Implement Tab 1 Radar View, fix background depth crash, and set main page redirect

feat: fix depth inversion and add video overlays

feat(inference): enable full multi-GPU support for all models

added depth view

added depth everything v2

fixed overflow

added apple depth pro

wired endpoints to new html

added async first frame/video detection