Vision-Language-Action models have recently emerged as a powerful paradigm for general-purpose robot learning, enabling agents to map visual observations and natural-language instructions into executable robotic actions. Though popular, they are primarily trained via supervised fine-tuning or training-time reinforcement learning, requiring explicit fine-tuning phases, human interventions, or controlled data collection. Consequently, existing methods remain unsuitable for challenging simulated- or physical-world deployments, where robots must respond autonomously and flexibly to evolving environments. To address this limitation, we introduce a Test-Time Reinforcement Learning for VLAs (TT-VLA), a framework that enables on-the-fly policy adaptation during inference. TT-VLA formulates a dense reward mechanism that leverages step-by-step task-progress signals to refine action policies during test time while preserving the SFT/RL-trained priors, making it an effective supplement to current VLA models. Empirical results show that our approach enhances overall adaptability, stability, and task success in dynamic, previously unseen scenarios under simulated and real-world settings. We believe TT-VLA offers a principled step toward self-improving, deployment-ready VLAs.

Generative models have been successfully applied in diverse domains, from natural language processing to image synthesis. However, despite this success, a key challenge that remains is the ability to control the semantic content of the scene being generated. We argue that adequate control of the generation process requires a data representation that allows users to access and efficiently manipulate the semantic factors shaping the data distribution. This work advocates for the adoption of succinct, informative, and interpretable representations, quantified using information-theoretic principles. Through extensive experiments, we demonstrate the efficacy of our proposed framework both qualitatively and quantitatively. Our work contributes to the ongoing quest to enhance both controllability and interpretability in the generation process. Code available at github.com/ArmandCom/InCoDe.

In this paper, the development of action and event detectors over the past three decades is summarized. The detectors are divided into 2D detectors, 3D detectors and deep learning detectors according to whether they contain spatial information and whether they use deep learning. This paper briefly introduces the typical detectors of the different types mentioned above, and explains the advantages, disadvantages and characteristics respectively, and compares them. Comparing traditional feature detection methods with ones based on deep learning, we found that the method of first detecting microscopic details such as point, line, surface angle, etc., and then performing action and event recognition is no longer the mainstream of current research. Due to the strong generalization ability, end-to-end action and event recognition methods based on deep learning perform better than traditional methods. Finally, this paper proposes three research directions for action recognition and event recognition based on feature detectors.

It is necessary to detect the fall of the elderly in time. As a noncontact monitoring device, radar can monitor users without their knowledge and protect their privacy. The unsupervised fall detection method does not need to collect and label fall samples, which avoids the difficulty of collecting fall data and saves researchers time and cost. The current unsupervised fall detection studies consider fewer types of actions and do not test the generalization of their models in new environments and subjects. This article proposes a new unsupervised fall detection system, including a feature extractor and predictor. We first use 3-D convolution and 3-D transposed convolution to construct a feature extractor to extract the range–velocity–time features of radar signals. Then, we construct a predictor to learn the pattern of nonfall action. Finally, we design an unsupervised training method based on hard sample mining to improve the ability of the model to identify hard negative samples. We train the model using only unlabeled nonfall samples and test it in new scenarios. The system’s accuracy in the data set containing 52 kinds of nonfall actions and 12 kinds of fall actions is 95.54%, the false alarm rate is 1.07%, and the area under the receiver operating characteristic is 0.9974.

Falls are fatal for the elderly, and timely detection after falls is crucial. As a contactless device, the radar sensor can monitor users’ falls with the advantage of not revealing their privacy. Today, the use of radar for fall detection has made a significant progress. However, most current methods cannot be used in real complex scenes. They usually collect fewer types of actions, and the ratio of the number of nonfall samples to the number of fall samples is small, which is not consistent with the real-life scene. In addition, the classifiers are usually trained and tested on the same environments and the same people, which cannot be easily extended to new environments and new people. We designed a robust fall detection system based on the frequency-modulated continuous-wave (FMCW) radar to solve these issues. The system detects the moment of human movement and calculates the range–velocity map, range–horizontal angle map, and range–vertical angle map of the radar signals, and creates three neural networks for these three signal maps. The stacking method of ensemble learning is used to fuse the time–space–velocity features extracted by the three neural networks to identify falls. The method was trained and tested on a data set consisting of ten scenes, 21 subjects, 52 nonfall action types, and 12 fall action types. The results show that on the test set containing only new environments and new subjects, the recall is 0.983 and the precision is 0.975.

Randomness and fuzziness are the two most important uncertainties inherent in human cognition, which have attracted great attention in artificial intelligence research. In this paper, regarding linguistic terms or concepts as the basic units of human cognition, we propose a new cognitive model—cloud model, which can synthetically describe the randomness and fuzziness of concepts and implement the uncertain transformation between a qualitative concept and its quantitative instantiations. Furthermore, by analyzing in detail the statistical properties of normal cloud model, that is, an important kind of cloud models based on normal distribution and Gauss membership function, we show that normal cloud model can not only be viewed as a generalized normal distribution with weak constraints but also avoid the flaw of fuzzy sets to quantify the membership degree of an element as an accurate value between 0 and 1 and, therefore, may be more adaptive for the uncertainty description of linguistic concepts. Finally, two demonstration examples about the fractal evolution of plants and network topologies based on cloud models are given to illustrate the promising applications of cloud models in some more complex knowledge representation tasks. © 2009 Wiley Periodicals, Inc.