Add exponentially longer sleeps to transcribe loop

When the user pauses their speech for an extended period of time, the
transcription engine will sleep for progressively longer intervals,
up to 1.5 seconds between transcriptions. This allows us to reduce
idle resource consumption.

To enable responsive transcription while the user is speaking actively,
we reset the sleep duration to the minimum whenever a change is
detected.

1 files changed, 2 insertions, 1 deletions

diff --git a/osc_ctrl.py b/osc_ctrl.py
index 4353939..328b55c 100644
--- a/osc_ctrl.py
+++ b/osc_ctrl.py
@@ -33,6 +33,8 @@ class EvilGlobalState():
     encoding = {}
 state = EvilGlobalState()
 
+# The characters in the TaSTT are all numbered from top left to bottom right.
+# This function provides a mapping from letter ('a') to index (26).
 def generateEncoding(state):
     for i in range(0, 26):
         state.encoding[chr(ord('A') + i)] = i
@@ -103,7 +105,6 @@ def sendMessageCellDiscrete(client, msg_cell, which_cell):
     if msg_cell != state.encoding[' '] * BOARD_COLS:
         addr="/avatar/parameters/" + generate_utils.getSpeechNoiseToggleParam()
         client.send_message(addr, False)
-        print("beep")
 
     time.sleep(CELL_TX_TIME_S / 3.0)