In the realm of robotics, numerous downstream robotics tasks leverage machine learning methods for processing, modeling, or synthesizing data. Often, this data comprises variables that inherently carry geometric constraints, such as the unit-norm condition of quaternions representing rigid-body orientations or the positive definiteness of stiffness and manipulability ellipsoids. Handling such geometric constraints effectively requires the incorporation of tools from differential geometry into the formulation of machine learning methods. In this context, Riemannian manifolds emerge as a powerful mathematical framework to handle such geometric constraints. Nevertheless, their recent adoption in robot learning has been largely characterized by a mathematically-flawed simplification, hereinafter referred to as the "single tangent space fallacy". This approach involves merely projecting the data of interest onto a single tangent (Euclidean) space, over which an off-the-shelf learning algorithm is applied. This paper provides a theoretical elucidation of various misconceptions surrounding this approach and offers experimental evidence of its shortcomings. Finally, it presents valuable insights to promote best practices when employing Riemannian geometry within robot learning applications.
相關內容
The effectiveness of reinforcement learning (RL) agents in continuous control robotics tasks is heavily dependent on the design of the underlying reward function. However, a misalignment between the reward function and user intentions, values, or social norms can be catastrophic in the real world. Current methods to mitigate this misalignment work by learning reward functions from human preferences; however, they inadvertently introduce a risk of reward overoptimization. In this work, we address this challenge by advocating for the adoption of regularized reward functions that more accurately mirror the intended behaviors. We propose a novel concept of reward regularization within the robotic RLHF (RL from Human Feedback) framework, which we refer to as \emph{agent preferences}. Our approach uniquely incorporates not just human feedback in the form of preferences but also considers the preferences of the RL agent itself during the reward function learning process. This dual consideration significantly mitigates the issue of reward function overoptimization in RL. We provide a theoretical justification for the proposed approach by formulating the robotic RLHF problem as a bilevel optimization problem. We demonstrate the efficiency of our algorithm {\ours} in several continuous control benchmarks including DeepMind Control Suite \cite{tassa2018deepmind} and MetaWorld \cite{yu2021metaworld} and high dimensional visual environments, with an improvement of more than 70\% in sample efficiency in comparison to current SOTA baselines. This showcases our approach's effectiveness in aligning reward functions with true behavioral intentions, setting a new benchmark in the field.
Integrating learning-based techniques, especially reinforcement learning, into robotics is promising for solving complex problems in unstructured environments. However, most existing approaches are trained in well-tuned simulators and subsequently deployed on real robots without online fine-tuning. In this setting, the simulation's realism seriously impacts the deployment's success rate. Instead, learning with real-world interaction data offers a promising alternative: not only eliminates the need for a fine-tuned simulator but also applies to a broader range of tasks where accurate modeling is unfeasible. One major problem for on-robot reinforcement learning is ensuring safety, as uncontrolled exploration can cause catastrophic damage to the robot or the environment. Indeed, safety specifications, often represented as constraints, can be complex and non-linear, making safety challenging to guarantee in learning systems. In this paper, we show how we can impose complex safety constraints on learning-based robotics systems in a principled manner, both from theoretical and practical points of view. Our approach is based on the concept of the Constraint Manifold, representing the set of safe robot configurations. Exploiting differential geometry techniques, i.e., the tangent space, we can construct a safe action space, allowing learning agents to sample arbitrary actions while ensuring safety. We demonstrate the method's effectiveness in a real-world Robot Air Hockey task, showing that our method can handle high-dimensional tasks with complex constraints. Videos of the real robot experiments are available on the project website (//puzeliu.github.io/TRO-ATACOM).
Automated data labeling techniques are crucial for accelerating the development of deep learning models, particularly in complex medical imaging applications. However, ensuring accuracy and efficiency remains challenging. This paper presents iterative refinement strategies for automated data labeling in facial landmark diagnosis to enhance accuracy and efficiency for deep learning models in medical applications, including dermatology, plastic surgery, and ophthalmology. Leveraging feedback mechanisms and advanced algorithms, our approach iteratively refines initial labels, reducing reliance on manual intervention while improving label quality. Through empirical evaluation and case studies, we demonstrate the effectiveness of our proposed strategies in deep learning tasks across medical imaging domains. Our results highlight the importance of iterative refinement in automated data labeling to enhance the capabilities of deep learning systems in medical imaging applications.
We propose a continuous-time nonlinear model of opinion dynamics with utility-maximizing agents connected via a social influence network. A distinguishing feature of the proposed model is the inclusion of an opinion-dependent resource-penalty term in the utilities, which limits the agents from holding opinions of large magnitude. The proposed utility functions also account for how the relative resources within the social group affect both an agent's stubbornness and social influence. Each agent myopically seeks to maximize its utility by revising its opinion in the gradient ascent direction of its utility function, thus leading to the proposed opinion dynamics. We show that, for any arbitrary social influence network, opinions are ultimately bounded. For networks with weak antagonistic relations, we show that there exists a globally exponentially stable equilibrium using contraction theory. We establish conditions for the existence of consensus equilibrium and analyze the relative dominance of the agents at consensus. We also conduct a game-theoretic analysis of the underlying opinion formation game, including on Nash equilibria and on prices of anarchy in terms of satisfaction ratios. Additionally, we also investigate the oscillatory behavior of opinions in a two-agent scenario. Finally, simulations illustrate our findings.
This study assesses deep learning models for audio classification in a clinical setting with the constraint of small datasets reflecting real-world prospective data collection. We analyze CNNs, including DenseNet and ConvNeXt, alongside transformer models like ViT, SWIN, and AST, and compare them against pre-trained audio models such as YAMNet and VGGish. Our method highlights the benefits of pre-training on large datasets before fine-tuning on specific clinical data. We prospectively collected two first-of-their-kind patient audio datasets from stroke patients. We investigated various preprocessing techniques, finding that RGB and grayscale spectrogram transformations affect model performance differently based on the priors they learn from pre-training. Our findings indicate CNNs can match or exceed transformer models in small dataset contexts, with DenseNet-Contrastive and AST models showing notable performance. This study highlights the significance of incremental marginal gains through model selection, pre-training, and preprocessing in sound classification; this offers valuable insights for clinical diagnostics that rely on audio classification.
Despite the recent increase in research activity, deep-learning models have not yet been widely accepted in several real-world settings, such as medicine. The shortage of high-quality annotated data often hinders the development of robust and generalizable models, which do not suffer from degraded effectiveness when presented with out-of-distribution (OOD) datasets. Contrastive Self-Supervised Learning (SSL) offers a potential solution to labeled data scarcity, as it takes advantage of unlabeled data to increase model effectiveness and robustness. However, the selection of appropriate transformations during the learning process is not a trivial task and even breaks down the ability of the network to extract meaningful information. In this research, we propose uncovering the optimal augmentations for applying contrastive learning in 1D phonocardiogram (PCG) classification. We perform an extensive comparative evaluation of a wide range of audio-based augmentations, evaluate models on multiple datasets across downstream tasks, and report on the impact of each augmentation. We demonstrate that depending on its training distribution, the effectiveness of a fully-supervised model can degrade up to 32%, while SSL models only lose up to 10% or even improve in some cases. We argue and experimentally demonstrate that, contrastive SSL pretraining can assist in providing robust classifiers which can generalize to unseen, OOD data, without relying on time- and labor-intensive annotation processes by medical experts. Furthermore, the proposed evaluation protocol sheds light on the most promising and appropriate augmentations for robust PCG signal processing, by calculating their effect size on model training. Finally, we provide researchers and practitioners with a roadmap towards producing robust models for PCG classification, in addition to an open-source codebase for developing novel approaches.
The pre-trained large language models (LLMs) have shown their extraordinary capacity to solve reasoning tasks, even on tasks that require a complex process involving multiple sub-steps. However, given the vast possible generation space of all the tasks, how the pretrained model learns the reasoning ability remains an open question. We firstly propose that an intrinsic structural constraint on the generated sequence of language-based reasoning -- we called it template-content structure (T-C structure) -- is the key to explain why LLMs can solve a large number of complex reasoning problems with limited training data by showing this structure can reduce the possible space from exponential level to linear level. Furthermore, by generalizing this structure to the hierarchical case, we demonstrate that models can achieve task composition, further reducing the space needed to learn from linear to logarithmic, thereby effectively learning on complex reasoning involving multiple steps. We provide both examples and formal theory of our T-C structure. We also experimentally validate the existence of the T-C structure in some current LLMs and its effectiveness for reasoning.
Ensuring the legal usage of deep models is crucial to promoting trustable, accountable, and responsible artificial intelligence innovation. Current passport-based methods that obfuscate model functionality for license-to-use and ownership verifications suffer from capacity and quality constraints, as they require retraining the owner model for new users. They are also vulnerable to advanced Expanded Residual Block ambiguity attacks. We propose Steganographic Passport, which uses an invertible steganographic network to decouple license-to-use from ownership verification by hiding the user's identity images into the owner-side passport and recovering them from their respective user-side passports. An irreversible and collision-resistant hash function is used to avoid exposing the owner-side passport from the derived user-side passports and increase the uniqueness of the model signature. To safeguard both the passport and model's weights against advanced ambiguity attacks, an activation-level obfuscation is proposed for the verification branch of the owner's model. By jointly training the verification and deployment branches, their weights become tightly coupled. The proposed method supports agile licensing of deep models by providing a strong ownership proof and license accountability without requiring a separate model retraining for the admission of every new user. Experiment results show that our Steganographic Passport outperforms other passport-based deep model protection methods in robustness against various known attacks.
STEMFold: Stochastic Temporal Manifold for Multi-Agent Interactions in the Presence of Hidden Agents
Learning accurate, data-driven predictive models for multiple interacting agents following unknown dynamics is crucial in many real-world physical and social systems. In many scenarios, dynamics prediction must be performed under incomplete observations, i.e., only a subset of agents are known and observable from a larger topological system while the behaviors of the unobserved agents and their interactions with the observed agents are not known. When only incomplete observations of a dynamical system are available, so that some states remain hidden, it is generally not possible to learn a closed-form model in these variables using either analytic or data-driven techniques. In this work, we propose STEMFold, a spatiotemporal attention-based generative model, to learn a stochastic manifold to predict the underlying unmeasured dynamics of the multi-agent system from observations of only visible agents. Our analytical results motivate STEMFold design using a spatiotemporal graph with time anchors to effectively map the observations of visible agents to a stochastic manifold with no prior information about interaction graph topology. We empirically evaluated our method on two simulations and two real-world datasets, where it outperformed existing networks in predicting complex multiagent interactions, even with many unobserved agents.
Denoising diffusion probabilistic models for image inpainting aim to add the noise to the texture of image during the forward process and recover masked regions with unmasked ones of the texture via the reverse denoising process. Despite the meaningful semantics generation, the existing arts suffer from the semantic discrepancy between masked and unmasked regions, since the semantically dense unmasked texture fails to be completely degraded while the masked regions turn to the pure noise in diffusion process, leading to the large discrepancy between them. In this paper, we aim to answer how unmasked semantics guide texture denoising process;together with how to tackle the semantic discrepancy, to facilitate the consistent and meaningful semantics generation. To this end, we propose a novel structure-guided diffusion model named StrDiffusion, to reformulate the conventional texture denoising process under structure guidance to derive a simplified denoising objective for image inpainting, while revealing: 1) the semantically sparse structure is beneficial to tackle semantic discrepancy in early stage, while dense texture generates reasonable semantics in late stage; 2) the semantics from unmasked regions essentially offer the time-dependent structure guidance for the texture denoising process, benefiting from the time-dependent sparsity of the structure semantics. For the denoising process, a structure-guided neural network is trained to estimate the simplified denoising objective by exploiting the consistency of the denoised structure between masked and unmasked regions. Besides, we devise an adaptive resampling strategy as a formal criterion as whether structure is competent to guide the texture denoising process, while regulate their semantic correlations. Extensive experiments validate the merits of StrDiffusion over the state-of-the-arts. Our code is available at //github.com/htyjers/StrDiffusion.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191