feat: replace SimpleTracker with ByteTrack
6896025
Zhen Ye
commited on
Spaces:
Sleeping
Sleeping
Commit History
Refine overlays: Hide on first frame and remove speed from labels
7a7588e
Zhen Ye
commited on
Respect CUDA_VISIBLE_DEVICES env var (do not delete it)
d3c7792
Zhen Ye
commited on
Fix NameError: import set_track_data in inference.py
03a0884
Zhen Ye
commited on
Add writer loop error logging
eb52404
Zhen Ye
commited on
Add aggressive worker debug logging and BaseException catch
6a62982
Zhen Ye
commited on
Add granular logging for frame processing and ffmpeg
bbef397
Zhen Ye
commited on
Fix backend worker pipeline stall on exception (add fallback fallback)
955fa9e
Zhen Ye
commited on
Fix video playback race condition, local file protocol, and backend worker crash
19fdabe
Zhen Ye
commited on
feat: Implement depth-based distance estimation and frontend integration
78d352c
Zhen Ye
commited on
fix: Increase queue buffer to absorb GPT latency
89b854c
Zhen Ye
commited on
feat: Real-time radar directionality via track sync
3a793fe
Zhen Ye
commited on
fix: Import Full/Empty from queue to prevent NameError on backpressure
90f0a6b
Zhen Ye
commited on
feat: Continuous object tracking, speed estimation, and overlay syncing
ff50694
Zhen Ye
commited on
Fix inference concurrency bugs and enable default first-frame GPT
06e44d3
Zhen Ye
commited on
fix: Move CUDA_VISIBLE_DEVICES clearing to module import time
9ea2cfe
Zhen Ye
commited on
perf: Tune batch sizes and queues for A10 GPUs
a5f8d15
Zhen Ye
commited on
feat: Implement AsyncVideoReader for parallel video decoding
356dce8
Zhen Ye
commited on
Uncomment CUDA_VISIBLE_DEVICES clearing to enable all GPUs
e43c504
Zhen Ye
commited on
Optimize streaming buffer and implement parallel depth pipeline
8d938e9
Zhen Ye
commited on
Refactor: Decouple depth estimator and suppress noisy logs
64bbe44
Zhen Ye
commited on
Implement Batch Inference & Queue Backpressure Fixes
5c36daa
Zhen Ye
commited on
Enable MJPEG streaming for live video feedback
a52a96d
Zhen Ye
commited on
fix: add wget to Dockerfile, cleanup js errors, and remove GPU env override
dc409c8
Zhen Ye
commited on
Integrated GPT Reasoning and Object Track Cards
fe2ee0d
Zhen Ye
commited on
Fix frontend layout and make depth estimation optional backend-wide
fbd1770
Zhen Ye
commited on
Feature: GPT distance integration, Radar Map, and optional legacy depth
0eeb0d9
Zhen Ye
commited on
Fix: GPU detection order and distance consistency
a69731b
Zhen Ye
commited on
fix:distance consistency
61fbbca
Zhen Ye
commited on
fix:multi-gpu support
fc53373
Zhen Ye
commited on
update:inference pipeline optimization
9606129
Zhen Ye
commited on
Fix GPU visibility, update depth scale, and improve radar display timing
be0e60f
Zhen Ye
commited on
Reduce log verbosity and update default depth scale to 100
ff62bd0
Zhen Ye
commited on
Optimize model loading to prevent reload loop
9a35d39
Zhen Ye
commited on
Fix draw_boxes arguments
8c16918
Zhen Ye
commited on
Fix IndentationError and undefined depth_raw variable in inference.py
9d9cecc
Zhen Ye
commited on
feat: fix depth inversion and add video overlays
391cb94
Zhen Ye
commited on
chore: add detailed GPU diagnostics logs
970cc85
Zhen Ye
commited on
feat: fix multi-gpu visibility and improve logging
0e0cc26
Zhen Ye
commited on
added debug level logging for GPU inferenciing and distance
4c39d24
Zhen Ye
commited on
feat(inference): enable full multi-GPU support for all models
45eb65b
Zhen Ye
commited on
added estimated/relative distance
1c2827d
Zhen Ye
commited on
added depth view
5e832fe
Zhen Ye
commited on
fixed overflow
78f99f1
Zhen Ye
commited on
fixed depth norm
6144c58
Zhen Ye
commited on
added fallback for Nan
91f3b56
Zhen Ye
commited on
added apple depth pro
b8fe2b6
Zhen Ye
commited on
wired endpoints to new html
9803004
Zhen Ye
commited on
fixed numpy error
3954f64
Zhen Ye
commited on
modified segmentation mask display
c7c9a25
Zhen Ye
commited on