![NXP Semiconductors SLN-LOCAL2-IOT User Manual Download Page 49](http://html.mh-extra.com/html/nxp-semiconductors/sln-local2-iot/sln-local2-iot_user-manual_1721901049.webp)
Table 10. Inference engine types (continued)
Inference engine type
asr_inference_t code
Encoding
Commands for Elevator
ASR_CMD_ELEVATOR
0x0004
Commands for Audio
Device Control
ASR_CMD_AUDIO
0x0008
Commands for Washing Machine
ASR_CMD_WASH
0x0010
Commands for LED Control
ASR_CMD_LED
0x0020
Commands for Dialog Stage 1
ASR_CMD_DIALOGIC_1
0x0040
Commands for Dialog Stage
2 Temperature
ASR_CMD_DIALOGIC_2_TEMPERATURE
0x0080
Commands for Dialog Stage
2 Timer
ASR_CMD_DIALOGIC_2_TIMER
0x0100
7.1.3 ASR configuration
7.1.3.1 Languages
The SLN-LOCAL2-IOT with i.MX RT106S can support up to four languages in runtime. If customized hardware or proof-of-
concepts are created, ensure the maximum number of languages to be enabled.
Figure 59. Configuration for the maximum number of languages snippet
• If i.MX RT105S is considered, it can support up to two languages. Set IMXRT105S to 1.
• If only one language is sufficient, we always recommend to set MULTILINGUAL to 0. This allows the ASR application to
save significant memory and CPU resources.
• Because MAX_CONCURRENT_LANGUAGES affects the resources, three microphones can be used when one or two
languages are enabled. Only two microphones are enabled for three or four languages.
7.1.3.2 Installation of languages and inference engines
Developers must ensure that the language models and inference engines are properly installed when initializing the ASR.
shows the language model installation function.
NXP Semiconductors
Far-field local voice control framework
SLN-LOCAL2-IOT Developer’s Guide, Rev. 0, 19 April 2021
User's Guide
49 / 87