Neural networks are information processing models based on the human brain, and they have been activity studied. However, in order to realize the hardware of the neural network, it is necessary to achieve high integration. In this study, we fabricated synapses for neural networks using the Ga-Sn-O (GTO) thin films, which is rare-metal-free amorphous oxide semiconductor. The synapses are planar type, and it is assumed that they are stacked on the LSI surface in the future. It was found that the synapses have a degradation characteristic that can be applied to modify Hebb's leaning rule. The deterioration characteristics obtained were modeled to simulate letter correction, and we succeeded the character correction.
We have developed a hybrid-type temperature sensor using thin-film transistors generating rectangle output waveform. The advantages of the hybrid-type temperature sensor are that the large temperature dependence of the off-leakage current can be utilized, and simultaneously only a digital circuit is required to count the digital pulse. However, the conventional hybrid-type sensor has a disadvantage that the pulse width is too small to count. Therefore, we have developed a new hybrid-type temperature sensor in order to enlarge the pulse width. Although we previously confirmed the operation by circuit simulation, particularly in this presentation, we have made an actual circuit and observed a rectangle output waveform, namely, confirmed the correct operation by actual experiment.
We have developed light irradiation and applied voltage history sensors using amorphous In-Ga-Zn-O (oc-IGZO) thin-film transistor (TFTs) exposed to ozone annealing and fabricated under high oxygen pressure. These a-IGZO TFTs have an interesting property; the Ids-Vgs characteristic shifts positively and becomes steep when gate voltage is applied whereas it recover the initial one with a small threshold voltage and a large subthreshold swing upon light illumination. Therefore, these a-IGZO TFTs can be used as light irradiation history sensors; first, the gate voltage is applied to initialize the Ids-Vgs characteristic, next, the light is irradiated and finally, the Ids-Vgs characteristic is measured which depends on the light irradiation history. Moreover, these a-IGZO TFTs can be also used as applied voltage history sensors; first, the light is irradiated to initialize the Ids-Vgs characteristic, next, the gate voltage is applied and finally, the Ids-Vgs characteristic is measured which depends on the value or time of the applied voltage.
In this presentation, we propose an amorphous oxide semiconductors (AOSs) In-Ga-Zn-O(IGZO) thin film for a memristor characteristic device. We fabricated the memristor characteristic device active layer using IGZO and electrodes using aluminum by physical vapor deposition (PVD). The AI/IGZO/Al cell device showed the bipolar switching characteristic of a switching voltage 2 and reproducibility 10.
Abstract We have developed infrared sensors using poly‐Si thin‐film transistors (TFTs) for proximity sensors integrated in smartphone displays. Initially, we evaluate the infrared sensitivities of the poly‐Si TFTs, and it is found that a pin‐type TFT is suitable for the infrared sensors. Next, we propose three types of the infrared sensors. First, an analog current detection‐type sensor has a simple structure, and it is found that it can detect presence of a hand. Second, a lock‐in detection‐type sensor has tolerance against ambient light, and it is found that it can detect a target signal under noise signals. Third, a frequency detection‐type sensor has an advantage that only a digital circuit is necessary for detection, and it is found that it can detect the infrared intensity because the oscillation frequency increases monotonically with the intensity. We can utilize these infrared sensors on demand.
Neural networks are information processing models based on the human brain, and they have been activity studied. However, in order to realize the hardware of the neural network, it is necessary to achieve high integration. In this study, we fabricated synapses for neural networks using the Ga-Sn-O (GTO) thin films, which is rare-metal-free amorphous oxide semiconductor. This synapse was a planar type and found to have degradation characteristics applicable to modified Hebb's learning rule. We used this result to deposit the GTO thin film on LSI surface. By considering the LSI as a neuron, it was possible to satisfy the relation between the synapse and the neuron, and as a result, a simple neural network could be constructed. Finally, we succeeded in character learning experiment using it.
Abstract We have confirmed working of a high‐resolutional active‐matrix flatpanel imager using poly‐Si thin‐film phototransistors for a magnifying viewer. The pixel resolution is 254 ppi, which is higher than the other ever‐developed ones. The flatpanel imager can correctly scan a font size of 1 mm, which is a sufficient reading ability.
