There is a growing demand for customized and expressive 3D characters with the emergence of AI agents and Metaverse, but creating 3D characters using traditional computer graphics tools is a complex and time-consuming task. To address these challenges, we propose a user-friendly framework named Make-A-Character (Mach) to create lifelike 3D avatars from text descriptions. The framework leverages the power of large language and vision models for textual intention understanding and intermediate image generation, followed by a series of human-oriented visual perception and 3D generation modules. Our system offers an intuitive approach for users to craft controllable, realistic, fully-realized 3D characters that meet their expectations within 2 minutes, while also enabling easy integration with existing CG pipeline for dynamic expressiveness. For more information, please visit the project page at https://human3daigc.github.io/MACH/.
With fewer feature dimensions, filter banks are often used in light-weight full-band speech enhancement models. In order to further enhance the coarse speech in the sub-band domain, it is necessary to apply a post-filtering for harmonic retrieval. The signal processing-based comb filters used in RNNoise and PercepNet have limited performance and may cause speech quality degradation due to inaccurate fundamental frequency estimation. To tackle this problem, we propose a learnable comb filter to enhance harmonics. Based on the sub-band model, we design a DNN-based fundamental frequency estimator to estimate the discrete fundamental frequencies and a comb filter for harmonic enhancement, which are trained via an end-to-end pattern. The experiments show the advantages of our proposed method over PecepNet and DeepFilterNet.
Target speaker information can be utilized in speech enhancement (SE) models to more effectively extract the desired speech. Previous works introduce the speaker embedding into speech enhancement models by means of concatenation or affine transformation. In this paper, we propose a speaker attentive module to calculate the attention scores between the speaker embedding and the intermediate features, which are used to rescale the features. By merging this module in the state-of-the-art SE model, we construct the personalized SE model for ICASSP Signal Processing Grand Challenge: DNS Challenge 5 (2023). Our system achieves a final score of 0.529 on the blind test set of track1 and 0.549 on track2.
We consider two biologically plausible structures, the Spiking Neural Network (SNN) and the self-attention mechanism. The former offers an energy-efficient and event-driven paradigm for deep learning, while the latter has the ability to capture feature dependencies, enabling Transformer to achieve good performance. It is intuitively promising to explore the marriage between them. In this paper, we consider leveraging both self-attention capability and biological properties of SNNs, and propose a novel Spiking Self Attention (SSA) as well as a powerful framework, named Spiking Transformer (Spikformer). The SSA mechanism in Spikformer models the sparse visual feature by using spike-form Query, Key, and Value without softmax. Since its computation is sparse and avoids multiplication, SSA is efficient and has low computational energy consumption. It is shown that Spikformer with SSA can outperform the state-of-the-art SNNs-like frameworks in image classification on both neuromorphic and static datasets. Spikformer (66.3M parameters) with comparable size to SEW-ResNet-152 (60.2M,69.26%) can achieve 74.81% top1 accuracy on ImageNet using 4 time steps, which is the state-of-the-art in directly trained SNNs models.
Lensless imaging is a popular research field for the advantages of small size, wide field-of-view and low aberration in recent years. However, some traditional lensless imaging methods suffer from slow convergence, mechanical errors and conjugate solution interference, which limit its further application and development. In this work, we proposed a lensless imaging method based on spatial light modulator (SLM) with unknown modulation curve. In our imaging system, we use SLM to modulate the wavefront of object, and introduce the ptychographic scanning algorithm that is able to recover the complex amplitude information even the SLM modulation curve is inaccurate or unknown. In addition, we also design a split-beam interference experiment to calibrate the modulation curve of SLM, and using the calibrated modulation function as the initial value of the expended ptychography iterative engine (ePIE) algorithm can improve the convergence speed. We further analyze the effect of modulation function, algorithm parameters and the characteristics of the coherent light source on the quality of reconstructed image. The simulated and real experiments show that the proposed method is superior to traditional mechanical scanning methods in terms of recovering speed and accuracy, with the recovering resolution up to 14 um.
Lacking enough high quality proposals for RoI box head has impeded two-stage and multi-stage object detectors for a long time, and many previous works try to solve it via improving RPN's performance or manually generating proposals from ground truth. However, these methods either need huge training and inference costs or bring little improvements. In this paper, we design a novel training method named APDI, which means augmenting proposals by the detector itself and can generate proposals with higher quality. Furthermore, APDI makes it possible to integrate IoU head into RoI box head. And it does not add any hyperparameter, which is beneficial for future research and downstream tasks. Extensive experiments on COCO dataset show that our method brings at least 2.7 AP improvements on Faster R-CNN with various backbones, and APDI can cooperate with advanced RPNs, such as GA-RPN and Cascade RPN, to obtain extra gains. Furthermore, it brings significant improvements on Cascade R-CNN.
Inertial navigation and attitude initialization in polar areas become a hot topic in recent years in the navigation community, as the widely-used navigation mechanization of the local level frame encounters the inherent singularity when the latitude approaches 90 degrees. Great endeavors have been devoted to devising novel navigation mechanizations such as the grid or transversal frames. This paper highlights the fact that the common Earth-frame mechanization is sufficiently good to well handle the singularity problem in polar areas. Simulation results are reported to demonstrate the singularity problem and the effectiveness of the Earth-frame mechanization.
Effective representation of a text is critical for various natural language processing tasks. For the particular task of Chinese sentiment analysis, it is important to understand and choose an effective representation of a text from different forms of Chinese representations such as word, character and pinyin. This paper presents a systematic study of the effect of these representations for Chinese sentiment analysis by proposing a multi-channel convolutional neural network (MCCNN), where each channel corresponds to a representation. Experimental results show that: (1) Word wins on the dataset of low OOV rate while character wins otherwise; (2) Using these representations in combination generally improves the performance; (3) The representations based on MCCNN outperform conventional ngram features using SVM; (4) The proposed MCCNN model achieves the competitive performance against the state-of-the-art model fastText for Chinese sentiment analysis.