Abstract: A robotic cleaner includes a housing, a suction conduit with an opening, and a leading roller mounted in front of a brush roll. An inter-roller air passageway may be defined between the leading roller and the brush roll wherein the lower portion of the leading roller is exposed to a flow path to the suction conduit and an upper portion of the leading roller is outside of the flow path. Optionally, a combing unit includes a plurality of combing protrusions extending into the leading roller and having leading edges not aligned with a center of the leading roller. Optionally, a sealing strip is located along a rear side of the opening and along a portion of left and right sides of the opening. The underside may define side edge vacuum passageways extending from the sides of the housing partially between the leading roller and the sealing strip towards the opening.

Abstract: In certain implementations, intent prediction is provided for a natural language utterance based on a portion of the natural language utterance prior to a system detection of an end of the natural language utterance. In some implementations, a first portion of a natural language utterance of a user may be received. Speech recognition may be performed on the first portion of the natural language utterance to recognize one or more words of the first portion of the natural language utterance. Context information for the natural language utterance may be obtained. Prior to a detection of an end of the natural language utterance, a first intent may be predicted based on the one or more words of the first portion and the context information. One or more user requests may be determined based on the first predicted intent.

Abstract: In certain implementations, a system and method for interpreting natural language inputs based on storage of the inputs is provided. A natural language input of a user may be obtained. The natural language input may be obtained via an input mode. The natural language input may be processed to determine a first interpretation of the natural language input. The natural language input may be stored based on a data format associated with the input mode. The natural language input may be obtained from storage. The natural language input obtained from storage may be reprocessed to determine a second interpretation of the natural language input.

Abstract: In certain implementations, intent prediction is provided for a natural language utterance based on a portion of the natural language utterance prior to a system detection of an end of the natural language utterance. In some implementations, a first portion of a natural language utterance of a user may be received. Speech recognition may be performed on the first portion of the natural language utterance to recognize one or more words of the first portion of the natural language utterance. Context information for the natural language utterance may be obtained. Prior to a detection of an end of the natural language utterance, a first intent may be predicted based on the one or more words of the first portion and the context information. One or more user requests may be determined based on the first predicted intent.

Abstract: In certain implementations, advertisers may be enabled to determine in real-time whether to include their advertisements (or other content) with reminders. As an example, the advertisement opportunity may be provided to the advertisers upon creation of the reminder. The advertisement opportunity may, for instance, comprise an opportunity to purchase an ad placement associated with the reminder. Upon determining that an advertiser purchased the associated ad placement, the associated ad placement may be allocated for an advertisement associated with the advertiser so that an advertisement associated with the advertiser may be presented with the reminder upon a triggering event indicating that the reminder should be presented to the user. When the reminder is provided for presentation to the user, a particular advertisement associated with the advertiser may be presented to the user.

Abstract: In certain implementations, a training environment is provided. A task that is to be performed is provided. The task may relate to a user request that is invokeable in one or more other environments different than the training environment. One or more user inputs may be received for the training environment. A determination of whether performance of the task has been satisfied may be effectuated based on the one or more user inputs. A reward may be provided for a user in response to a determination that performance of the task has been satisfied. In some implementations, the determination of whether performance of the task has been satisfied may comprise a determination of whether the user request would have been invoked in the one or more other environments if the user inputs had been received for the one or more other environments.

Abstract: A protective bank damming device useful for damming and controlling the flow of a liquid, the liquid most commonly being water. The damming device barrier member is inflated with water to provide a form to the barrier member, and anchoring weight to keep the barrier secured to the substrate surface. The barrier member provides support for a skirt member extending from the top side of the barrier to a skirt distal edge abutting against the flood side substrate surface. A skirt sealing means is applied along the skirt distal edge, with the purpose of preventing leakage under the skirt and barrier member, and utilizing natural leakage inherent in this type of liquid-anchored dam, reducing the hydrostatic pressure under the skirt and barrier relative to the dammed water. This pressure differential keeps the damming device securely anchored to the surface.