Abstract: Provided is an image recognition method including the steps of: for a deep learning network that carries out object recognition on a random image, carrying out quantization corresponding to the number of a plurality of different bits to generate a plurality of quantization models respectively corresponding to the number of bits; receiving image data as an input for the deep learning network; determining the uncertainty of the input image data; selecting any one of the plurality of quantization models on the basis of the determined uncertainty; and recognizing an object from the image data by using the selected quantization model, and outputting, as the result of the object recognition, a label corresponding to the image data.

Abstract: An image recognition method includes the steps of: for a deep learning network that carries out object recognition on a random image, carrying out quantization corresponding to the number of a plurality of different bits to generate a plurality of quantization models respectively corresponding to the number of bits; receiving image data as an input for the deep learning network; determining the uncertainty of the input image data; selecting any one of the plurality of quantization models on the basis of the determined uncertainty; and recognizing an object from the image data by using the selected quantization model, and outputting, as the result of the object recognition, a label corresponding to the image data.

Abstract: Provided are a pixel processing apparatus, and an image signal processing apparatus and an image processing system each including the same. The pixel processing apparatus includes a bad pixel processing unit configured to determine whether a target pixel is a bad pixel based on pixel data of an image sensor, output a bad pixel detection signal according to a result of determining whether the target pixel is a bad pixel, compensate for a pixel value of the target pixel based on the pixel data, and output the compensated pixel value; and a noise removal unit configured to remove noise from the pixel value of the target pixel, based on the bad pixel detection signal, the pixel value, the compensated pixel value, and weight information.

Abstract: An image signal processor including a first data processing circuit, a second data processing circuit, and a multiplexing control circuit configured to transmit first line data included in first frame data and second line data included in second frame data to the first data processing circuit in a time-division multiplexing mode may be provided. The multiplexing control circuit may be further configured to transmit to the first data processing circuit a switching control signal, which indicates whether current line data to be transmitted to the first data processing circuit is the first line data or the second line data.

Abstract: An image signal processor including a first data processing circuit, a second data processing circuit, and a multiplexing control circuit configured to transmit first line data included in first frame data and second line data included in second frame data to the first data processing circuit in a time-division multiplexing mode may be provided. The multiplexing control circuit may be further configured to transmit to the first data processing circuit a switching control signal, which indicates whether current line data to be transmitted to the first data processing circuit is the first line data or the second line data.

Abstract: Provided are a pixel processing apparatus, and an image signal processing apparatus and an image processing system each including the same. The pixel processing apparatus includes a bad pixel processing unit configured to determine whether a target pixel is a bad pixel based on pixel data of an image sensor, output a bad pixel detection signal according to a result of determining whether the target pixel is a bad pixel, compensate for a pixel value of the target pixel based on the pixel data, and output the compensated pixel value; and a noise removal unit configured to remove noise from the pixel value of the target pixel, based on the bad pixel detection signal, the pixel value, the compensated pixel value, and weight information.