Table 1 Comparison of in-sensor computing devices based on conventional and 2D materials.
From: Two-dimensional material-based devices for in-sensor computing
Materials | Sensory source | Application | Energy consumption (J) | Operating voltage (V) | Endurance (cycles) | Retention (s) | Light wavelength (nm) |
|---|---|---|---|---|---|---|---|
Conventional materials | |||||||
ZnO/Ag/ITO [197] | Light | Face recognition | 1.0 × 10−6 | 2 | 500 | 1.0 × 104 | white |
Si [42] | Light | Edge computing | 1.0 × 10−7 | 2.5 | >1 × 103 | – | 650–950 |
CuPc/P(VDF-TrFE) [199] | Light | Pain formation simulation | 4.6 × 10−8 | 0.5 | - | >5.0 × 101 | 660, 445 |
ITO/SA Biopolymer [198] | Light | Pain perception | 10−8 to 10−7 | 2 | >250 | - | 360–860 |
2D materials | |||||||
MoS2/BTO [50] | Light | Digit recognition | 1.8 × 10−12 | 5 | 100 | 1.0 × 105 | 450–650 |
Gr/MoTe2/P(VDF-TrFE) [38] | Light | Edge detection | 1.0 × 10−13 | 15 | 1 × 106 | 9.0 × 104 | 340–1310 |
α-In2Se3 (2H) [56] | Light | Lane keeping | 2.7 × 10−7 | 3 | 1 × 104 | 1.7 × 105 | 532 |
Black phosphorus [87] | Light | Pavlovian learning | 9.2 × 10−10 | 0.01 | – | – | 280–365 |
VO2 (20 nm film) [93] | Light | Image recognition | 4.0 × 10−8 | 2 | – | 4.0 × 103 | 375–532 |
MXene–ZnO [86] | Humidity & Light | Image recognition | 4.5 × 10−6 | 3 | – | 1.0 × 104 | 365 |
VO2 (40 nm film) [19] | Temperature & pressure | human machine interaction | 3.9 × 10−9 | 2 | 1 × 1012 | - | - |
Ti3C2Tx MXenes [85] | Light | Supervised learning | 5 × 10−15 | 0.05 | >550 | >200 | 365 |
MoSSe [32] | Light | Light adaptation | 4 × 10−9 | 4 | >4000 | >100 | 350–550 |
MoS2/h-BN/Te [94] | Light | Dimensionality reduction | 3.2 × 10−10 | 5 | 1 × 104 | >1.0 × 103 | 450–635 |
