FocsleFo’c’sle
GO

EfficientDet-Lite4

4%
by Google

Reference EfficientDet variant pinned at INT8 for legacy NPUs. Maintained, not advanced.

Object detectionApache-2.0INT8cocolegacy
312K downloads 11K deploymentsUpdated Nov 8, 2027
Headline:19.2ms · Google Coral Edge TPU · INT8
OverviewBenchmarks6Sim ResultsDeploy6FilesDiscussion23

HIL simulation runs

End-to-end runs of EfficientDet-Lite4 in scenario rigs that match real customer conditions. Each tile is an actual playable HIL session — sustained throughput, in-sim accuracy, intervention counts, the whole trace.

Queue a new run
car · 0.94truck · 0.87car · 0.78
48 FPS94.1% acc
Open
AV · Highway

Av highway · EfficientDet-Lite4

EfficientDet-Lite4×GGoogle Coral Edge TPU
37 min runNov 8, 2027
car · 0.94truck · 0.87car · 0.78
24 FPS88.6% acc
Open
AV · Urban

Av urban · EfficientDet-Lite4

EfficientDet-Lite4×QQualcomm QCS6490
70 min runNov 8, 2027
car · 0.94truck · 0.87car · 0.78
18 FPS90.4% acc
Open
Robotics · AMR

Warehouse amr · EfficientDet-Lite4

EfficientDet-Lite4×AAmbarella CV5
61 min runNov 8, 2027
car · 0.94truck · 0.87car · 0.78
58 FPS92.8% acc
Open
Logistics · Apron

Airport baggage · EfficientDet-Lite4

EfficientDet-Lite4×πRaspberry Pi 5 + Hailo HAT
32 min runNov 8, 2027
car · 0.94truck · 0.87car · 0.78

Pro tier opens up scenario authoring

Pro and Team customers can configure custom HIL scenarios — chain CARLA towns, drop in your own actor policies, and pin the exact downstream consumers your fleet runs. Enterprise extends this to dedicated rigs in the lab and on-prem replication.

See plansHow the lab works →