close
RTI to GoDownload RTI APP now
Open
:::

WATCH: AI helps temples prevent fire and decode handwriting

  • 27 December, 2024
  • Amanda Ruth Stephens
WATCH: AI helps temples prevent fire and decode handwriting
This isn't the first time the Wude Temple in Beigang has employed AI-- in July of this year, the temple introduced an AI system to help people know how to read poems in 30 seconds, which reduced pressure on staff and improved overall efficiency. (Photo: CNA)

While Wude Temple’s AI application was originally meant to just count visitors, it has been expanded to include an additional smoke and fire alert system, and to decode illegible handwriting.

Video script: 

Tendrils of smoke curl through the temple air, carrying visitors' prayers to heaven. However, the hot smoke and burning incense also pose a fire risk– a problem some temples are turning to AI to solve.

According to Wude Temple Secretary-General Su Chien-hua (蘇建華), AI technology assesses the volume of smoke and warns workers to inspect potential danger sites based on the amount of smoke noticed. 

Though originally designed to count the number of visitors over a month, or even a year, the AI system is proving to be both more accurate than previous systems and also capable of identifying worshippers’ age, and gender. 

But it doesn’t just help identify devotees, it can also help them decode the sometimes messy handwriting of the temple masters. The temple masters often write back interpretations in response to questions, but sometimes the words are too scrawled to be read easily by anyone than a fellow master. That’s where the system comes in.

The AI system can help identify characters, making decoding advice from the temple even more efficient– especially considering that the temple houses hundreds of thousands of tomes filled with heavenly advice. And temple officials aren’t waiting around for it to go to waste– in the future they plan to cooperate with universities to turn Taiwan’s intangible beliefs from something only read in smoke into a database read on the web.

Amanda Stephens, for Rti News

Comments

Latest Newsmore