Abstract: A visual content search and retrieval platform comprising an object detection model, a coarse-grained image classification model, and a selection of fine-grained image classification models. During inference, the object detection model locates and extracts a query object in a query image; the coarse-grained image classification model classifies the query object's meta-category; the fine-grained image classification model trained and optimized for the meta-category of the query object is selected; and the selected fine-grained image classification model searches and retrieves from an image database images of objects most similar to the query object.

Abstract: A display device includes a driving substrate, a plurality of light-emitting elements, a seed layer, a metal layer, and an insulating layer. The plurality of light-emitting elements are electrically connected to the driving substrate. The seed layer is disposed on the driving substrate and located between the plurality of light-emitting elements. The metal layer is disposed on the seed layer, wherein the metal layer surrounds the light-emitting elements. The insulating layer surrounds the light-emitting element and contacts the metal layer and a sidewall of the light-emitting elements respectively.

Abstract: An analysis method and an analysis device for a semiconductor manufacturing process are provided. The analysis method of the semiconductor manufacturing process includes the following steps. A semiconductor machine log file is loaded. A plurality of machine action events are defined from the semiconductor manufacturing machine log file. A wafer moving path map is generated according to a wafer moving information file. A plurality of account point information are bound to each of the machine action events according to the wafer moving path map. A wafer movement tracking graph is generated according to the machine action events that have been bound to account information.

Abstract: An electro-acoustical transducer device is disclosed, which includes: a hollow disk body that generally defines an axis of propagation, the hollow disk body comprising: a pair of plate members extending substantially perpendicular to the axis of propagation, each provided with a central transmitting port arranged about the axis of propagation, and a peripheral enclosure jointing the pair of plate members at the respective outer edge portions thereof, thereby defining a chamber of resonance between the pair of plate members; wherein a ring-opening about the axis of propagation that enables access to the chamber of resonance is formed between the central transmitting ports of the plate members.

Abstract: A control device is configured to control a display panel according to a trigger signal transmitted from an input device to a host. The control device includes a connector unit and a signal capture unit. The connector unit is configured to receive the trigger signal, and transmit the same to the host. The signal capture unit is configured to capture the trigger signal, and control the display panel according to the trigger signal while the connector unit is transmitting the trigger signal to the host.

Abstract: An input device is provided to supply multiple input from a button. The input device may include a first switch and a second switch. The input device may further include a slide button on a surface of the input device. The slide button may actuate the first switch when the slide button slides along a first direction substantially parallel to the surface of the input device. The slide button may actuate the second switch when the slide button slides along a second direction substantially parallel to the surface of the input device. The input device may further include a third switch. The slide button may actuate the third switch of the input device when the slide button is pressed towards the surface of the input device.

Abstract: A computerized method of automatic distributed communication includes training a first and second machine learning models with historical feature vector inputs to generate a likelihood output and a mean count output, respectively. For each entity in a set, the method includes processing a likelihood feature vector input with the first machine learning model to generate a likelihood output indicative of a likelihood that the entity will have an avoidable negative health event within a specified first time period, and processing a mean count feature vector input with the second machine learning model to generate a mean count output indicative of an expected number of avoidable negative health events that the entity will have within a specified second time period. The method includes automatically distributing structured campaign data to at least a subset of entities in the set according to the likelihood output or the mean count output.

Abstract: A computer-implemented method for efficiently encoding video may include (i) determining, by a computing device, that a video file is a candidate for vertical-slice-based region-of-interest compression, (ii) identifying a video encoder that supports horizontal-slice-based compression but does not support vertical-slice-based compression, (iii) rotating each frame of the video file ninety degrees, and (iv) performing, by the video encoder, horizontal-slice-based region-of-interest compression on the rotated video file. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: An electroluminescent device, wherein the electroluminescent device includes a first-conductivity-type semiconductor layer, a second-conductivity-type semiconductor layer, an active layer, a first electrode, a second electrode, and an optical conversion material. The active layer is disposed between the first-conductivity-type semiconductor layer and the second-conductivity-type semiconductor layer and electrically connected with these two. The first-conductivity-type semiconductor layer has a light-emitting surface disposed on a side opposite to the active layer, and includes a plurality of 3D structures arranged regularly, extending from the light-emitting surface towards the active layer to jointly define at least one cavity having a depth greater than 70% a thickness of the first-conductivity-type semiconductor layer. The optical conversion material is filled in the cavity.

Abstract: A method includes receiving an input text utterance to be synthesized into expressive speech having an intended prosody and a target voice and generating, using a first text-to-speech (TTS) model, an intermediate synthesized speech representation for the input text utterance. The intermediate synthesized speech representation possesses the intended prosody. The method also includes providing the intermediate synthesized speech representation to a second TTS model that includes an encoder portion and a decoder portion. The encoder portion is configured to encode the intermediate synthesized speech representation into an utterance embedding that specifies the intended prosody. The decoder portion is configured to process the input text utterance and the utterance embedding to generate an output audio signal of expressive speech that has the intended prosody specified by the utterance embedding and speaker characteristics of the target voice.

Abstract: The present invention discloses methods and systems for providing UICC/eUICC related response information to information requests at a cellular router. The method includes receiving an information request from a wireless communication module, and determining whether a response to the information request is cached. When the response information is not cached, forwarding the information request to a massive SIM apparatus (MSA). MSA will then respond to the information request. A response based on the MSA's response will then be sent to the wireless communication module for the information request. When the response information is cached, retrieve the response information and send it to the wireless communication module.

Abstract: An input device is provided to supply multiple input from a button. The input device may include a first switch and a second switch. The input device may further include a slide button on a surface of the input device. The slide button may actuate the first switch when the slide button slides along a first direction substantially parallel to the surface of the input device. The slide button may actuate the second switch when the slide button slides along a second direction substantially parallel to the surface of the input device. The input device may further include a third switch. The slide button may actuate the third switch of the input device when the slide button is pressed towards the surface of the input device.

Abstract: An image processing apparatus includes an output circuit, a processing circuit and a receiving circuit. The output circuit includes connection ports, and is configured to perform handshakes through connected ones of connection ports to respectively obtain display capability data through connected ones of connection ports. The output circuit is further configured to output picture data through connected ones of connection ports, respectively. The processing circuit, coupled with the output circuit, is configured to determine a stitching mode resolution according to display capability data. The receiving circuit, coupled with the processing circuit, is configured to perform handshakes according to the stitching mode resolution, in order to obtain a first image having the stitching mode resolution. The processing circuit is configured to slice the first image into sub-images corresponding to the display capability data, and to generate picture data including the sub-images.

Abstract: A cross-shape transfer blade is configured to, in operation, be mounted or coupled to an end of a transfer robot arm (TRA). The cross-shape robot or transfer blade includes a plurality of raised regions to contact a backside of a workpiece or thin workpiece such that the cross-shape transfer blade supports and transports the respective workpeice between various locations within a semiconductor manufacturing plant (FAB). The cross-shape transfer blade includes a first prong structure, a second prong structure, a first wing structure, and a second wing structure. Respective ones of the plurality of raised regions are at corresponding ones of the first prong structure, the second prong structure, the first wing structure, and the second wing structure minimizing contact between the backside of the workpiece or thin workpiece and the cross-shape transfer blade when the workpiece or thin workpiece is being transferred, transported, and supported by the cross-shape transfer blade.

Abstract: An analog to digital converter includes voltage divider circuits, front-end circuits, at least one converter circuit, and a controller circuit. The voltage divider circuits are configured to divide an input signal to generate first signals, in which the first signals have different levels. The front-end circuits are configured to respectively sample the first signals to generate second signals. The at least one converter circuit is configured to generate at least one digital output according to the second signals and a reference voltage. The controller circuit is configured to determine a level of the input signal according to the at least one digital output and select one of the at least one digital output according to the level of the input signal.

Abstract: A method and a system for automatic evaluation of online debate arguments are provided. The method includes: defining multiple persona knowledge dimensions; tailor-making a prompt template according to the defined persona knowledge dimensions to generate a persona knowledge eliciting prompt; inputting the persona knowledge eliciting prompt to a pretrained language model to elicit persona knowledge; associating the elicited persona knowledge and the context and argument in the prompt to construct a knowledge-aligned prompt; and utilizing the knowledge-aligned prompt to fine-tune a debate evaluation model to evaluate persuasiveness/impact of the online debate arguments.

Abstract: A method for representing an intended prosody in synthesized speech includes receiving a text utterance having at least one word, and selecting an utterance embedding for the text utterance. Each word in the text utterance has at least one syllable and each syllable has at least one phoneme. The utterance embedding represents an intended prosody. For each syllable, using the selected utterance embedding, the method also includes: predicting a duration of the syllable by decoding a prosodic syllable embedding for the syllable based on attention by an attention mechanism to linguistic features of each phoneme of the syllable and generating a plurality of fixed-length predicted frames based on the predicted duration for the syllable.

Abstract: An automatic cooking system with a cooking tool includes an ingredient storage and feeding apparatus and an ingredient conveying and dispensing apparatus. The ingredient storage and feeding apparatus includes a plurality of storage units with inlets and outlets, a plurality of ingredient boxes are provided between the inlets and outlets. The ingredient conveying and dispensing apparatus includes a work position located on one side of the cooking tool. The ingredient conveying and dispensing apparatus transports the ingredient boxes to the work position and then rotates so that the ingredient boxes are turned over to pour the required ingredients into the cooking tool. The automatic cooking system realizes the integration of the classified storage and transportation of ingredients, a single ingredient is used as a storage unit, and ingredients can be automatically transported from different storage units to the pot according to the order of adding the ingredients to the dishes.

Abstract: A display light engine which includes a light source module, a spectrum conversion array disposed on the light source module and a lens array is provided. The light source module includes a plurality of light-emitting components which emit light toward the spectrum conversion array. The spectrum conversion array includes a plurality of spectrum conversion components and a metal bank material. The spectrum conversion components are spaced from each other and are disposed on the light-emitting components respectively. The metal bank material is distributed between two adjacent spectrum conversion components and separates the spectrum conversion components from each other. The lens array which includes a plurality of lenses arranged in an array is disposed between the spectrum conversion array and the light source module. Those lenses are disposed on and align to the spectrum conversion components respectively.

Abstract: A method includes obtaining training data including a plurality of training audio signals and corresponding transcripts. Each training audio signal is spoken by a target speaker in a first accent/dialect. For each training audio signal of the training data, the method includes generating a training synthesized speech representation spoken by the target speaker in a second accent/dialect different than the first accent/dialect and training a text-to-speech (TTS) system based on the corresponding transcript and the training synthesized speech representation. The method also includes receiving an input text utterance to be synthesized into speech in the second accent/dialect. The method also includes obtaining conditioning inputs that include a speaker embedding and an accent/dialect identifier that identifies the second accent/dialect.