Abstract: Provided is a method of generating a language model using crossmodal information. The method includes: receiving language-based first modality information and non-language-based second modality information; converting the first modality information into a first byte sequence; converting the second modality information into a second byte sequence; converting the first and second byte sequences into a first embedding vector and a second embedding vector by applying an embedding technique for each modality; generating semantic association information between first and second modality information by inputting the first and second embedding vectors to a crossmodal transformer; and learning the language model by setting the generated semantic association information as training data.

Abstract: A method of controlling a towing mode of an eco-friendly vehicle, may easily charge a battery of a towed vehicle by selecting one of a charge mode during traveling, a regenerative braking charge mode, and a coasting charge mode for charging the battery of the towed vehicle when a towing vehicle tows the towed vehicle so that a motor of the towed vehicle is driven by a negative charge torque for charging the battery when the towing vehicle accelerates or decelerates, and furthermore, may easily perform the assist of the traveling driving force of the towing vehicle by selecting a discharge mode during traveling in a state in which the towed vehicle does not need to be charged to the amount of charge of the battery of the towed vehicle greater than or equal to a reference value so that the motor of the towed vehicle is driven by a positive driving torque by discharging the battery of the towed vehicle.

Abstract: Image sensors include an array of image sensor pixels therein. This array of image sensor pixels includes a first focus detection pixel and at least a first color pixel. A switching network is provided, which is electrically coupled to the array. This switching network may be configured to generate a first mixed image signal by electronically mixing a focus detection signal generated by the first focus detection pixel with at least one color pixel signal generated by the at least a first color pixel. The first focus detection pixel can be a color-blind pixel, which may include a light-blocking shield mask therein.

Abstract: Image sensors include an array of image sensor pixels therein. This array of image sensor pixels includes a first focus detection pixel and at least a first color pixel. A switching network is provided, which is electrically coupled to the array. This switching network may be configured to generate a first mixed image signal by electronically mixing a focus detection signal generated by the first focus detection pixel with at least one color pixel signal generated by the at least a first color pixel. The first focus detection pixel can be a color-blind pixel, which may include a light-blocking shield mask therein.