Tomohiro Mashita

Kazufumi Hosoda, Keigo Nishida, Shigeto Seno et al.

Learning a new concept from one example is a superior function of the human brain and it is drawing attention in the field of machine learning as a one-shot learning task. In this paper, we propose one of the simplest methods for this task with a nonparametric weight imprinting, named Direct ONE-shot learning (DONE). DONE adds new classes to a pretrained deep neural network (DNN) classifier with neither training optimization nor pretrained-DNN modification. DONE is inspired by Hebbian theory and directly uses the neural activity input of the final dense layer obtained from data that belongs to the new additional class as the synaptic weight with a newly-provided-output neuron for the new class, transforming all statistical properties of the neural activity into those of synaptic weight by quantile normalization. DONE requires just one inference for learning a new concept and its procedure is simple, deterministic, not requiring parameter tuning and hyperparameters. DONE overcomes a severe problem of existing weight imprinting methods that DNN-dependently interfere with the classification of original-class images. The performance of DONE depends entirely on the pretrained DNN model used as a backbone model, and we confirmed that DONE with current well-trained backbone models perform at a decent accuracy.

Xuanmeng Sha, Liyun Zhang, Tomohiro Mashita et al.

Generating holistic co-speech gestures that integrate full-body motion with facial expressions suffers from semantically incoherent coordination on body motion and spatially unstable meaningless movements due to existing part-decomposed or frame-level regression methods, We introduce 3DGesPolicy, a novel action-based framework that reformulates holistic gesture generation as a continuous trajectory control problem through diffusion policy from robotics. By modeling frame-to-frame variations as unified holistic actions, our method effectively learns inter-frame holistic gesture motion patterns and ensures both spatially and semantically coherent movement trajectories that adhere to realistic motion manifolds. To further bridge the gap in expressive alignment, we propose a Gesture-Audio-Phoneme (GAP) fusion module that can deeply integrate and refine multi-modal signals, ensuring structured and fine-grained alignment between speech semantics, body motion, and facial expressions. Extensive quantitative and qualitative experiments on the BEAT2 dataset demonstrate the effectiveness of our 3DGesPolicy across other state-of-the-art methods in generating natural, expressive, and highly speech-aligned holistic gestures.

Xuanmeng Sha, Liyun Zhang, Tomohiro Mashita et al.

Audio-driven 3D facial animation has achieved significant progress in both research and applications. While recent baselines struggle to generate natural and continuous facial movements due to their frame-by-frame vertex generation approach, we propose 3DFacePolicy, a pioneer work that introduces a novel definition of vertex trajectory changes across consecutive frames through the concept of "action". By predicting action sequences for each vertex that encode frame-to-frame movements, we reformulate vertex generation approach into an action-based control paradigm. Specifically, we leverage a robotic control mechanism, diffusion policy, to predict action sequences conditioned on both audio and vertex states. Extensive experiments on VOCASET and BIWI datasets demonstrate that our approach significantly outperforms state-of-the-art methods and is particularly expert in dynamic, expressive and naturally smooth facial animations.

Sota Shoman, Tomohiro Mashita, Alexander Plopski et al.

Accurate camera localization is an essential part of tracking systems. However, localization results are greatly affected by illumination. Including data collected under various lighting conditions can improve the robustness of the localization algorithm to lighting variation. However, this is very tedious and time consuming. By using synthesized images it is possible to easily accumulate a large variety of views under varying illumination and weather conditions. Despite continuously improving processing power and rendering algorithms, synthesized images do not perfectly match real images of the same scene, i.e. there exists a gap between real and synthesized images that also affects the accuracy of camera localization. To reduce the impact of this gap, we introduce "REal-to-Synthetic Transform (REST)." REST is an autoencoder-like network that converts real features to their synthetic counterpart. The converted features can then be matched against the accumulated database for robust camera localization. In our experiments REST improved feature matching accuracy under variable lighting conditions by approximately 30%. Moreover, our system outperforms state of the art CNN-based camera localization methods trained with synthetic images. We believe our method could be used to initialize local tracking and to simplify data accumulation for lighting robust localization.