Original link: https://cdjax.com/?p=571
In retrospect, I experienced IBM’s voice input for the first time on a PC around 2002. Through the ViaVoice software, I could type in Word by speaking with a headset. This was my first encounter with “artificial intelligence”.
But now, we are completely accustomed to speech-to-text, copying text from pictures, etc., so natural that it is like breathing.
When it comes to language translation, the progress over the years has also been remarkable. Machine translation today is smoother and more natural than it was ten years ago. Especially after combining image recognition technology, we can directly take pictures for translation, and the effect is also greatly improved. Driven by large-scale language models such as the GPT series, the effect of machine translation is even close to the level of native speakers.
For driving, the high-computing chip and the cameras all over the body have improved the vehicle’s own perception capabilities and made our driving safer. In many scenarios, assisted driving has been more stable than myself. For example, the human eye can no longer see the lane markings clearly in rainy days, but the 3D picture presented by computer vision shows whether I am still in the lane in milliseconds.
Besides these, many more possibilities are being explored. For example, we can use image recognition technology to identify and track animals to help us better protect the ecological environment. We can also use speech recognition technology to help hearing-impaired people understand the words of others, so that they can better participate in social interactions.
The applications of these technologies are limitless, they make our lives more convenient and provide us with more opportunities and possibilities. We look forward to more surprises and changes they can bring us in the future. And we, together with these technologies, will continue to explore and advance to create a better future.
This article is transferred from: https://cdjax.com/?p=571
This site is only for collection, and the copyright belongs to the original author.