trash image classification - Impulses

Beta / trash image classification Public

Name	Input	DSP Blocks	Learn Blocks	Tc	Lr	Do	Ne	i8_V_Acc	i8_T_Acc	Tr_Proc	i8_Latency	i8_RAM	i8_Flash
Impulse #1	96×96	Image (RGB)	Classifier	40	0.001	-	-	70.1%	-	CPU	1053 ms.	186.3K	45.5K	View impulse
Impulse #2	96×96	Image (RGB)	Classifier	55	0.0005	-	-	72.2%	55.3%	CPU	872 ms.	184.3K	41.5K	View impulse
best model	160×160	Image (RGB)	MobileNetV2 160x160 0.35	20	0.001	0.1	0	85.7%	-	CPU	3104 ms.	416.1K	650.6K	View impulse
Impulse #6	160×160	Image (RGB)	MobileNetV2 160x160 0.35	50	0.0005	0.2	8	82.5%	75.8%	CPU	4236 ms.	416.2K	536.5K	View impulse
Impulse #7	96×96	Image (RGB)	MobileNetV2 96x96 0.05	60	0.001	0.1	8	65.3%	-	CPU	1114 ms.	159.3K	150.8K	View impulse
Impulse #10	96×96	Image (RGB)	MobileNetV2 96x96 0.1	60	0.001	0.1	0	75.2%	-	CPU	1284 ms.	169.1K	231.1K	View impulse
Impulse #11	160×160	Image (RGB)	Transfer learning	-	-	-	-	84.4%	-	CPU	4299 ms.	416.1K	650.6K	View impulse
Impulse #12	160×160	Image (RGB)	MobileNetV2 160x160 0.35	50	0.001	0.2	0	85.0%	83.3%	CPU	4688 ms.	416.1K	650.6K	View impulse
Impulse #13	96×96	Image (Grayscale)	MobileNetV2 96x96 0.35	80	0.0001	0.1	4	78.1%	68.6%	CPU	1648 ms.	214.6K	530.6K	View impulse
best model with mod	160×160	Image (RGB)	Transfer learning	-	-	-	-	86.0%	80.6%	CPU	4249 ms.	416.1K	650.6K	View impulse