Abstract: Disclosed herein is a system and method for improving the accuracy of an object detector when trained with a dataset having a significant number of missing annotations. The method uses a novel Background Recalibration Loss (BRL) which adjusts the gradient direction according to its own activation to reduce the adverse effect of error signals by replacing the negative branch of the focal loss with a mirror of the positive branch when the activation is below a confusion threshold.

Abstract: A system and method of fine-grained image category discovery with few human annotations includes a camera and a trained machine learning model, which predicts a label for an object in a captured image and outputs the predicted label. The machine learning model is trained by contrastive affinity learning, including retrieving images having an object, a warm-up stage in which semi-supervised contrastive learning is performed based on projected features of a class token and an ensembled prompt, respectively. In a contrastive affinity learning stage, a student model and an exponentially moving averaged teacher model are forwarded with different augmented views of the retrieved images. Teacher embeddings are enqueued into a token-specific memory. A semi-supervised contrastive loss is computed on a current batch and a contrastive affinity learning loss for student embeddings and the teacher embeddings with pseudo-labels from a affinity graph dynamically generated by semi-supervised affinity generation.

Abstract: Disclosed herein is a system and method for novel neural architecture search using a random graph network backbone to facilitate the creation of an efficient network structure. The method utilizes reinforcement learning algorithms to build a complex relationship between intra-connections (i.e., links between blocks in a random graph network) and extra-connections (i.e., links among blocks across the random graphs network) for discovering an efficient random neural architecture.

Abstract: Disclosed herein is an improvement to prior art feature pyramids for general object detection that inserts a simple norm calibration (NC) operation between the feature pyramids and detection head to alleviate and balance the norm bias caused by feature pyramid network (FPN) and which leverages an enhanced multi-feature selective strategy (MS) during training to assign the ground-truth to one or more levels of the feature pyramid.

Abstract: Disclosed herein is a system and method for novel neural architecture search using a random graph network backbone to facilitate the creation of an efficient network structure. The method utilizes reinforcement learning algorithms to build a complex relationship between intra-connections (i.e., links between blocks in a random graph network) and extra-connections (i.e., links among blocks across the random graphs network) for discovering an efficient random neural architecture.

Abstract: Disclosed herein is a method of soft anchor-point detection (SAPD), which implements a concise, single-stage anchor-point detector with both faster speed and higher accuracy. Also disclosed is a novel training strategy with two softened optimization techniques: soft-weighted anchor points and soft-selected pyramid levels.

Abstract: Disclosed herein an effective detach strategy which suppresses the flow of gradients from context sub-networks through the detection backbone path to obtain a more discriminative context by forcing the representation of context sub-network to be dissimilar from the detection network. A sub-network is defined to generate the context information from early layers of the detection backbone. Because instance and context focus on perceptually different parts of an image, the representations from either of them should also be discrepant. In addition, a stacked complementary loss is generated to and backpropagated to the detection network.

Abstract: An image generation method includes: extracting a content feature of a first image; respectively extracting an overall image style feature of a second image, and an object style feature of a partial image block, which includes an object, in the second image; determining a target style feature at least according to the overall image style feature and the object style feature; and generating a third image according to the content feature and the target style feature.

Abstract: The present disclosure relates to a blowing structure and a fan. The blowing structure includes a main body, a base, a control panel, a second drive member, and a first drive member. The main body is provided with a blowing channel, and the main body is rotatably connected to the base. The control panel is disposed on the main body. The second drive member is disposed in the blowing channel. The first drive member is connected to the main body, and the first drive member is configured to adjust an angle of rotation of the main body with respect to the base. The control panel is electrically connected to the second drive member and the first drive member.

Abstract: Disclosed herein is a design for a 1-bit CNN that closes the performance gap between binary neural networks and real-valued networks on challenging large-scale datasets. The design starts with a high-performance baseline network. Blocks with identity shortcuts which bypass 1-bit generic convolutions are adopted to replace the convolutions in the baseline network. Reshaping and shifting of activation functions is introduced. Finally, a distributional loss to further is adopted enforce the binary network to learn similar output distributions as those of a real-valued network.

Abstract: Disclosed herein is a method providing a flexible way to transfer knowledge from base to novel classes in a few shot learning scenario. The invention introduces a partial transfer paradigm for the few-shot classification task in which a model is first trained on the base classes. Then, instead of transferring the learned representation by freezing the whole backbone network, an efficient evolutionary search method is used to automatically determine which layer or layers need to be frozen and which will be fine-tuned on the support set of the novel class.

Abstract: Techniques and apparatus for generating dense natural language descriptions for video content are described. In one embodiment, for example, an apparatus may include at least one memory and logic, at least a portion of the logic comprised in hardware coupled to the at least one memory, the logic to receive a source video comprising a plurality of frames, determine a plurality of regions for each of the plurality of frames, generate at least one region-sequence connecting the determined plurality of regions, apply a language model to the at least one region-sequence to generate description information comprising a description of at least a portion of content of the source video. Other embodiments are described and claimed.

Abstract: Disclosed herein is a system and method for improving the accuracy of an object detector when trained with a dataset having a significant number of missing annotations. The method uses a novel Background Recalibration Loss (BRL) which adjusts the gradient direction according to its own activation to reduce the adverse effect of error signals by replacing the negative branch of the focal loss with a mirror of the positive branch when the activation is below a confusion threshold.

Abstract: Techniques and apparatus for generating dense natural language descriptions for video content are described. In one embodiment, for example, an apparatus may include at least one memory and logic, at least a portion of the logic comprised in hardware coupled to the at least one memory, the logic to receive a source video comprising a plurality of frames, determine a plurality of regions for each of the plurality of frames, generate at least one region-sequence connecting the determined plurality of regions, apply a language model to the at least one region-sequence to generate description information comprising a description of at least a portion of content of the source video. Other embodiments are described and claimed.

Abstract: Techniques and apparatus for generating dense natural language descriptions for video content are described. In one embodiment, for example, an apparatus may include at least one memory and logic, at least a portion of the logic comprised in hardware coupled to the at least one memory, the logic to receive a source video comprising a plurality of frames, determine a plurality of regions for each of the plurality of frames, generate at least one region-sequence connecting the determined plurality of regions, apply a language model to the at least one region-sequence to generate description information comprising a description of at least a portion of content of the source video. Other embodiments are described and claimed.

Abstract: An image generation method includes: extracting a content feature of a first image; respectively extracting an overall image style feature of a second image, and an object style feature of a partial image block, which includes an object, in the second image; determining a target style feature at least according to the overall image style feature and the object style feature; and generating a third image according to the content feature and the target style feature.

Abstract: Disclosed herein is an improvement to prior art feature pyramids for general object detection that inserts a simple norm calibration (NC) operation between the feature pyramids and detection head to alleviate and balance the norm bias caused by feature pyramid network (FPN) and which leverages an enhanced multi-feature selective strategy (MS) during training to assign the ground-truth to one or more levels of the feature pyramid.

Abstract: Techniques and apparatus for generating dense natural language descriptions for video content are described. In one embodiment, for example, an apparatus may include at least one memory and logic, at least a portion of the logic comprised in hardware coupled to the at least one memory, the logic to receive a source video comprising a plurality of frames, determine a plurality of regions for each of the plurality of frames, generate at least one region-sequence connecting the determined plurality of regions, apply a language model to the at least one region-sequence to generate description information comprising a description of at least a portion of content of the source video. Other embodiments are described and claimed.