Question at position 9 _________ requires lots of processing power and is therefore mainly used with Internet-enabled devices today.CAPTCHAHandwriting recognitionOptical mark recognitionSpeech recognition单项选择题
A
CAPTCHA
B
Handwriting recognition
C
Optical mark recognition
D
Speech recognition

Question

Question at position 9 _________ requires lots of processing power and is therefore mainly used with Internet-enabled devices today.CAPTCHAHandwriting recognitionOptical mark recognitionSpeech recognition单项选择题

A

CAPTCHA

B

Handwriting recognition

C

Optical mark recognition

D

Speech recognition

BlackTom AI · Accepted Answer

Speech recognition systems analyze continuous audio streams, extract acoustic features, run large statistical or neural models, and search language models — all of which demand significant processing power and large datasets, which is why they are mainly used with Internet-enabled devices today.

1. Relevant knowledge points and concepts
- Speech recognition: converts spoken language to text using signal processing, feature extraction (MFCCs), acoustic models, and language models (often deep neural networks), requiring high compute for training and often for inference.
- CAPTCHA: simple image/text tasks designed to distinguish humans from bots; low processing needs.
- Handwriting recognition: optical character recognition for cursive/printed writing; moderate compute, often localized.
- Optical mark recognition (OMR): detects filled bubbles/marks; minimal processing.

2. Step-by-step reasoning
- Audio processing: continuous audio sampled at e.g. 16 kHz produces many frames per second; each frame requires feature computation.
- Model complexity: modern speech systems use deep nets with millions of parameters; real-time decoding requires beam search across hypotheses and language model scoring.
- Latency and robustness: noise reduction and adaptation add extra processing. Many devices offload heavy tasks to cloud servers via Internet.

3. Intermediate-detail considerations (no numeric algebra needed)
- Compare compute profiles: OMR: simple thresholding over a few pixels -> tiny compute. CAPTCHA: basic image rendering and simple verification -> low. Handwriting recognition: per-character classification with smaller models -> moderate. Speech recognition: continuous time-series + large models -> highest.

4. Why the correct answer is correct
Speech recognition uniquely combines continuous real-time signal processing with large statistical models, creating the largest processing burden among the options; thus it is predominantly provided via Internet-enabled services or cloud processing.

5. Why other options are incorrect
- CAPTCHA: lightweight, not compute-intensive.
- Handwriting recognition: moderate complexity but less than continuous speech and often runs locally.
- Optical mark recognition: simplest, minimal processing.

6. Conclusion
The correct answer is Speech recognition, matching the provided answer.

Question at position 9 _________ requires lots of processing power and is therefore mainly used with Internet-enabled devices today.CAPTCHAHandwriting recognitionOptical mark recognitionSpeech recognition单项选择题

类似问题

The technology that converts speech signals into digital data and allows user to commend the system to perform tasks or ask questions is known as_____.

Which AI frontier enables machines to understand and respond to human speech?

10. Speech-to-text software can be specially designed to help deal with Answer Question 8[input].

Identify the feature outlined in the red dashed circle in this image.

Which of the following is true of the use of female rodents in basic research?

Which of the following is true about sperm competition and male physiology?

Rick presented an experiment in Lecture 3.3 ("Are You Experienced?") about the impacts of cocaine on sexual behaviour. What did the authors find?

According to findings from a mate choice experiment discussed in the Week 1 Lecture "Different Strokes...", female rats show a decreased preference for ________________.

更多留学生实用工具

智能学习助手

风格化写作助手

论文查重助手

文献引用助手

课堂转译助手

课堂笔记助手

Quiz搜索助手

学校历年真题

智能刷题助手

智能匹配练习

希望你的学习变得更简单