Skip to main content

Table 3 Main features of the most relevant user identification works surveyed

From: Continuous user identification in distance learning: a recent technology perspective

Work

ID type

Core method

Offline

Computation cost

ID precision

Needed hardware

Belhumeur et al. (1997)

Image-based

LDA

✓

\(\star\)

\(\star\)

Webcam

He et al. (2005)

Image-based

LPP + Nearest-neighbor classifier

✓

\(\star \star\)

\(\star \star\)

Webcam

Ahonen et al. (2006)

Image-based

LBP + Bayesian classifier

✓

\(\star \star\)

\(\star \star \star\)

Webcam

Heisele et al. (2001)

Image-based

SVM

✓

\(\star \star\)

\(\star \star \star\)

Webcam

Cao et al. (2018)

Image-based

ResNet50 CNN (DL)

✓

\(\star \star \star\)

\(\star \star \star \,\star\)

Webcam+GPU

Taigman et al. (DeepFace) (2014)

Image-based

DeepFace CNN (DL)

✓

\(\star \star \star \star \star\)

\(\star \star \star \star \star\)

Webcam+GPU

Taigman et al. (FaceNet) (2014)

Image-based

Inception CNN (DL) + Distance threshold

✓

\(\star \star \star \,\star\)

\(\star \star \star \star \star\)

Webcam+GPU

VGGFace Parkhi et al. (2015)

Image-based

VGGFace CNN (DL)

✓

\(\star \star \star \,\star\)

\(\star \star \star \star \star\)

Webcam+GPU

Stolcke et al. (2007)

Voice-based

MLLR transforms + SVM

✓

\(\star \star \star\)

\(\star \star \star\)

Microphone

Seurin et al. (2020)

Voice-based

MDP + RL

✓

\(\star \star \star\)

\(\star \star \star \,\star\)

Microphone

Boles and Rad (2017)

Voice-based

MFCCs + SVM NN (DL)

✗

\(\star \star \star \star \star\)

\(\star \star \star \star \star\)

Microphone

Ravanelli and Bengio (2018)

Voice-based

Band-Pass Filters + CNN (DL)

✓

\(\star \star \star \,\star\)

\(\star \star \star \,\star\)

Microphone

Nagrani et al. (2020)

Voice-based

Relation Module + Two-stream synchronization CNN (DL)

✓

\(\star \star \star \star \star\)

\(\star \star \star \star \star\)

Microphone

Bergadano et al. (2002)

Interaction-based

Keystroke trigraph duration

✓

\(\star\)

\(\star \star \star\)

Keyboard

Clarke and Furnell (2007)

Interaction-based

FF MLP Neural Network

✓

\(\star \star \star \,\star\)

\(\star \star \star\)

Mobile PhoneHandset

Shen et al. (2015)

Interaction-based

Feature-distance vectors + SVM

✓

\(\star \star\)

\(\star \star \star\)

Touch-basedSmartphone

Zheng et al. (2016)

Interaction-based

Mouse angle-based metrics + SVM

✓✗

\(\star\)

\(\star \star \star\)

Mouse

Chang et al. (2012)

Password and interaction-based

Graphical password + Touch time and pressure statistical classifier

✓

\(\star\)

\(\star \star \star \,\star\)

Touch-based Mobile Device

  1. Computation cost and Identification (ID) precision rated as Very Low (\(\star\)), Low (\(\star \star\)), Medium (\(\star \star \star\)), High (\(\star \star \star \,\star\)) or Very High (\(\star \star \star \star \star\))
  2. CNN: Convolutional neural network
  3. DL: Deep learning
  4. FF MLP: Feed forward multi-layered perceptron
  5. LDA: Linear discriminant analysis
  6. LBP: Local binary patterns
  7. LPP: Locality preserving projection
  8. MDP: Markov decision process
  9. MFCC: Mel-frequency cepstral coefficient
  10. MLLR: Maximum likelihood linear regression
  11. RL: Reinforcement learning
  12. SVM: Support vector machine