PressureLens - Web

Video: loading | OCR: loading | Level: lightregister pressure model

data logging:

samples: 0 | voice: 0 | selected topics: 0 | page topics: no

pressure model user:pressure model: final13 base

pressure bar

light (0-33%)medium (33-66%)hard (66-100%)

input mode:

intention detection:

realtime speed: 0.0 px/s

finger long-press LLM:no auto OCR/LLM on finger hold

pressure request LLM:No Hand · pressure request: idle

pressure level:

response mode:

image enhancement:(raw camera image)

warp:

00.50.18

fingertip comp:

00.120.057

perspective:

0%100%80%

(near large, far small effect)

Model Input

180x180 RGB

Pressure

Idle

No Hand

Firm0%

Light0%

NoPress0%

Model

final13_all_subjects_base_crop180.onnx

Patch180px to 180px

Latency--

Response

Tap the video to OCR the region under your pen, then call LLM.

topics & voice notestap to expand / collapse

Topics (for recommender)

no topics yet - run OCR first

Press and speak (save as topic)

voice topic:

wake word:for always listen mode

speech sensitivity:0.015