Abstract: In a signal transmission method and apparatus, a network device sends scheduling information of at least one of two carriers whose communication directions are opposite on a target OFDM symbol; a terminal device performs channel transmission of a target carrier on the target OFDM symbol based on the scheduling information, where the target carrier is one of carriers that are of the two carriers whose communication directions are opposite and that have channel transmission on the target OFDM symbol; and the network device performs channel transmission on the target OFDM symbol and the target carrier. A time domain resource may be allocated to a carrier on which transmission scheduling is currently performed, thereby improving transmission resource utilization and reducing a waiting delay.

Abstract: A communication method and apparatus. for indicating indicate A terminal is indicated to perform second transmission when an SSB or a PRACH on a frequency domain resource overlaps with reverse transmission on another frequency domain resource in time domain. In response to the SSB or the PRACH overlapping with the second transmission in time domain, the terminal device determines, by receiving first indication information, that a priority of the SSB or the PRACH is lower than a priority of the second transmission on an overlapped time domain resource. The terminal device performs the second transmission based on the first indication information. The terminal device performs reverse transmission on a symbol that overlaps with the SSB or the PRACH.

Abstract: An electronic device may be provided with peripheral conductive housing structures, a first antenna, and a second antenna. A gap may divide the housing structures into a first segment forming an arm of the first antenna and a second segment forming an arm of the second antenna. A first feed terminal may be coupled to the first segment and a second feed terminal may be coupled to the second segment. Switchable components may be coupled in parallel between the first and second feed terminals across the gap. The switchable components may be adjusted to tune the frequency response of the first and/or second antenna. The switchable components may have a first state in which only the first feed terminal feeds the first antenna and may have a second state in which both the first and second feed terminals feed the first antenna.

Abstract: The apresnt disclosure adopts the thermal bonding process to process the microchannel heat sink. By placing the upper cover plate and the lower cover plate on the plates of the microchannel heat sink, the pressure is directly applied, and there is no need to add other adhesives.

Abstract: A method is provided that includes training a CLIP model to learn embeddings of images and text from matched image-text pairs. The text represents image attributes. The method trains a StyleGAN on images in a training dataset of matched image-text pairs. The method also trains, using a CLIP model guided contrastive loss which attracts matched text embedding pairs and repels unmatched pairs, a text-to-direction model to predict a text direction that is semantically aligned with an input text responsive to the input text and a random latent code. A triplet loss is used to learn text directions using the embeddings learned by the trained CLIP model. The method generates, by the trained StyleGAN, positive and negative synthesized images by respectively adding and subtracting the text direction in the latent space of the trained StyleGAN corresponding to a word for each of the words in the training dataset.

Abstract: In one example, a computing device includes at least one processor configured to output for display, a graphical keyboard. The at least one processor may also be configured to model, in a first lattice, a plurality of candidate character strings that include symbols of a first alphabet. The first lattice may indicate spatial probabilities of the plurality of candidate character strings. The at least one processor may be configured to determine, using a second lattice that indicates probabilities of one or more words of a second language based at least in part on the spatial probabilities of the plurality of candidate character strings, a probability that the at least one of the plurality of candidate character strings corresponds to at least one word included in the second language. The at least one processor may be configured to output for display, the one or more symbols representing at least one word.

Abstract: Methods and systems for inputting tone and diacritic marks by gesture are provided. A method may include displaying a keypad for text input. Selection of a key and a gesture may be received. The gesture input may be semantically linked to a phonological property, a diacritic property, and/or a tone. A gesture input may include a substantially horizontal movement to the right associated with a first tone, a movement up and to the right associated with a second tone, a movement down and to the right, then up and to the right associated with a third tone, and a movement down and to the right associated with a fourth tone. A grapheme may be displayed based upon the selected key and the received gesture. With the diacritic and/or tone information, a grapheme may be predicted which may help to reduce the effort and speed of input.

Abstract: In one example, a computing device includes at least one processor configured to output for display, a graphical keyboard. The at least one processor may also be configured to model, in a first lattice, a plurality of candidate character strings that include symbols of a first alphabet. The first lattice may indicate spatial probabilities of the plurality of candidate character strings. The at least one processor may be configured to determine, using a second lattice that indicates probabilities of one or more words of a second language based at least in part on the spatial probabilities of the plurality of candidate character strings, a probability that the at least one of the plurality of candidate character strings corresponds to at least one word included in the second language. The at least one processor may be configured to output for display, the one or more symbols representing at least one word.

Abstract: Methods and systems for inputting tone and diacritic marks by gesture are provided. A method may include displaying a keypad for text input. Selection of a key and a gesture may be received. The gesture input may be semantically linked to a phonological property, a diacritic property, and/or a tone. A gesture input may include a substantially horizontal movement to the right associated with a first tone, a movement up and to the right associated with a second tone, a movement down and to the right, then up and to the right associated with a third tone, and a movement down and to the right associated with a fourth tone. A grapheme may be displayed based upon the selected key and the received gesture. With the diacritic and/or tone information, a grapheme may be predicted which may help to reduce the effort and speed of input.