Counter Audio Capture & Analytics
Directional Array Pickup
On-Prem Live ASR
Local Privacy Redaction
Scenario details
For fixed counters such as retail checkouts, bank and public service windows, pharmacy desks, and consultation tables. The three tabs below cover what the on-site AI hardware does.
Clear Pickup in Noise
Pain Points & Challenges
Bank counters, public service windows, and retail checkout lanes suffer background chatter, keyboard clicks, and queue chimes; basic microphones record distorted audio that severely degrades downstream speech-to-text accuracy.
Edge AI Solution
Deploy reSpeaker 4/6-mic arrays on counters with built-in AEC and noise reduction for 3–5m directional pickup and DOA < 5° error, suppressing lateral noise to focus on counter dialogue.
Core Advantages
- On-Chip DSP: reSpeaker built-in AEC and noise suppression filter ambient interference
- Directional: 4/6-mic array delivers 3–5m directional pickup with DOA < 5°
- Plug & Play: Standard UAC interface delivers clean 16kHz audio over USB
Echo & Noise Suppression
Directional Pickup
Plug-and-Play USB
Retail Voice
Ready to deployLocal Live Transcription
Pain Points & Challenges
Branch outlets generate thousands of daily customer dialogue recordings; streaming all audio to headquarters clouds consumes heavy WAN bandwidth and exposes sensitive IDs, account numbers, and passwords to compliance hazards.
Edge AI Solution
Deploy a low-power reComputer edge host in branch cabinets to run SenseVoice ASR (TensorRT / RKNN fp16) with 167ms~583ms typical latency, redacting sensitive data locally and uploading only structured text.
Core Advantages
- Live ASR: SenseVoice transcribes on-prem with 167ms~583ms typical latency per sentence
- Audio Privacy: Raw audio streams never leave the device; speech is redacted on-prem
- Offline Resync: Sentences cache locally during outages and resume on reconnect
Edge Transcription
Text-Only Upload
Keeps Recording Offline
Retail Voice
Ready to deploySpeaker Separation
Pain Points & Challenges
Single-channel recordings blend employee and customer voices into an inseparable stream; automated systems cannot tell who uttered which phrase, rendering employee compliance checks and service rating rules unusable.
Edge AI Solution
Combine reSpeaker mic arrays and reComputer edge hosts with diarization algorithms and DOA < 5° localization, segmenting timestamps within 20ms frames into structured turn-taking dialogue.
Core Advantages
- Role Separation: Combines DOA < 5° localization and diarization to parse dialogue
- Voiceprint Enrollment: Enroll staff voiceprints locally on reComputer with 3–5 phrases
- Privacy & Security: Raw audio and voiceprint vectors remain on the edge host
Speaker per Sentence
Staff Voiceprints
Voiceprints Stay Local
Retail Voice
Ready to deployDeployment & Selection Guide
The configurator narrows it down to a model by transcription mode, pickup setup and compute; the two architectures and device comparisons are in the cards below.
Not sure which to choose? See the full selection guide →
Transcription Mode
Configure Your Solution
Choose transcription mode and collection devices as needed to generate a recommended list.
Cloud Transcribe Kit
reSpeaker collection → network → SenseCraft Voice transcription → analytics dashboard
硬件/物料名称 数量 型号/SKU 说明 reSpeaker XVF3800 with Case (cased, plug-and-play on the counter) 1 114993701 - Local Transcribe Kit
reSpeaker collection → local AI box ASR → optional cloud sync
硬件/物料名称 数量 型号/SKU 说明 reSpeaker XVF3800 with Case (cased, plug-and-play on the counter) 1 114993701 - reRouter CM4 (entry-level local transcription) 1 110110205 -
Deployment Architecture
Cloud gateway relay or offline edge ASR
Cloud transcription streams audio to cloud via gateway without local edge box; on-premise mode runs offline ASR on edge box with audio kept local. Both setups share the same front-end mic arrays.
Core Advantages
- Cloud Pipeline: Lightweight gateway streams audio without on-site compute
- Offline Privacy: Edge host transcribes locally so audio never leaves premises
- Shared Hardware: Both deployment modes use identical front-end mic arrays
| Image | Product | Tier | Capabilities | Measured Streams | Scale |
|---|---|---|---|---|---|
 | reRouter CM4 1432 | Entry | Single-language local transcription | 1–2 streams (CPU) | Small single-point deployments, validated reference design |
 | reComputer AI R2130-12 | Advanced | Real-time multilingual transcription | 6 streams (p95 1173 ms) | Medium scale, latency-sensitive |
 | reComputer J4012 | High throughput | Real-time multilingual transcription | 48 streams (p95 789 ms) | Many concurrent streams on one box |
 | reComputer J3011 | High throughput | Real-time multilingual transcription | 32 streams (p95 293 ms) | Many concurrent streams on one box |
 | reComputer RK3576-30 | Standalone | Multilingual transcription + local analysis | 12 streams (p95 1067 ms) | High privacy or offline environments |
 | reComputer RK3588-30 | Standalone | Multilingual transcription + local analysis | 8 streams (p95 902 ms) | High privacy or offline environments |
Local Transcription Hosts
Edge hosts run offline ASR with zero leaks
Edge hosts run offline transcription on-premise so audio never leaves the store. Match compute tiers by stream count and offline analysis needs; see measured throughput on the left.
Core Advantages
- High Throughput: Single box handles up to 48 concurrent live streams
- On-Device Masking: Sensitive fields are redacted on site with zero leakage
- Turnkey Setup: Validated official reference design ready for deployment
Retail Voice
Ready to deploy| Image | Product | Type | Range | Coverage | Amp | Algorithms |
|---|---|---|---|---|---|---|
 | reSpeaker Lite | Linear dual-mic | ~3m | 180° | 5W | AEC · DoA |
 | reSpeaker XVF3800 | Circular quad-mic | ~5m | 360° | 5W | AEC · DoA · Multi-beam |
 | reSpeaker XVF3800 (with Case) | Circular quad-mic | ~5m | 360° | 5W | AEC · AGC · DoA · VAD · Multi-beam |
 | reSpeaker Flex Circular-4 | Circular quad-mic | ~5m | 360° | 10W | AEC · DoA · Multi-beam |
 | reSpeaker Flex Linear-4 | Linear quad-mic | ~5m | 180° | 10W | AEC · DoA · Multi-beam |
Audio Capture Devices
Select mic arrays by distance and coverage
Pick directional arrays for 1-on-1 counters; circular arrays for service desks; and Flex for 10W amp or custom enclosures.
* For wearable mobile badges, see In-Person Conversation Intelligence.
Core Advantages
- Directional Dual-Mic: Beamforming suppresses side noise within three meters
- Omni Quad-Mic: Circular four-mic array covers 360 degrees at service desks
- Preset Acoustics: Pre-tuned DSP algorithms plug and play over USB
Frequently Asked Questions
Yes. Seeed provides ODM/OEM customization services that can integrate microphone arrays and edge AI modules into a single hardware design, suitable for customers who need unified procurement or branded deployment. If you have customization needs, feel free to reach out: Ask about ODM/OEM customization →
Yes. SenseCraft Voice provides open APIs, and transcription results can be pushed in real time through standard interfaces to CRM, ticketing, ERP, and other business systems. No manual export is required, and custom data formats and callback endpoints are supported.
Use the wearable reSpeaker Clip — 360° pickup at about 3 m, 14–18 hours of recording without a phone nearby, and audio uploaded through the phone app for transcription. See In-Person Conversation Intelligence for the wearable deployment.
Yes. The Seeed devices used in this solution can be customized with your own logo, enclosure, packaging and pre-loaded firmware, and existing models can be adapted to add or drop interfaces and features. See Customization Service for scope and process, or .