灰度图像着色是AI在信息恢复中的引人入胜的应用。该问题的天生性质不良的性质使其更具挑战性，因为输出可能是多模式的。目前正在使用的基于学习的方法为直接情况产生可接受的结果，但在没有明确的图形分离的情况下通常无法恢复上下文信息。同样，由于在完整图像特征上训练的单个模型不足以学习各种数据模式，因此图像遭受了颜色出血和饱和背景。为了解决这些问题，我们提出了一个基于GAN的配色框架。在我们的方法中，每个量身定制的GAN管道都会使前景（使用对象级特征）或背景（使用全图像功能）着色。前景管道采用了一个具有自我注意事项的残留无UNET作为其发电机，使用了全图像功能和可可数据集中的相应对象级特征训练。背景管道依赖于该位置数据集的全图像功能和其他培训示例。我们设计了一个基于密集的融合网络，以通过基于特征的融合来获得最终的有色图像。我们显示了通常用于评估多模式问题（例如图像着色）并使用多个感知指标对我们的框架进行广泛的绩效评估的非感知评估指标的缺点。我们的方法的表现优于大多数基于学习的方法，并且产生的结果与最新的方法相当。此外，我们进行了运行时分析，并获得了每个图像的平均推理时间24ms。

Model compression via quantization and sparsity enhancement has gained an immense interest to enable the deployment of deep neural networks (DNNs) in resource-constrained edge environments. Although these techniques have shown promising results in reducing the energy, latency and memory requirements of the DNNs, their performance in non-ideal real-world settings (such as in the presence of hardware faults) is yet to be completely understood. In this paper, we investigate the impact of bit-flip and stuck-at faults on activation-sparse quantized DNNs (QDNNs). We show that a high level of activation sparsity comes at the cost of larger vulnerability to faults. For instance, activation-sparse QDNNs exhibit up to 17.32% lower accuracy than the standard QDNNs. We also establish that one of the major cause of the degraded accuracy is sharper minima in the loss landscape for activation-sparse QDNNs, which makes them more sensitive to perturbations in the weight values due to faults. Based on this observation, we propose the mitigation of the impact of faults by employing a sharpness-aware quantization (SAQ) training scheme. The activation-sparse and standard QDNNs trained with SAQ have up to 36.71% and 24.76% higher inference accuracy, respectively compared to their conventionally trained equivalents. Moreover, we show that SAQ-trained activation-sparse QDNNs show better accuracy in faulty settings than standard QDNNs trained conventionally. Thus the proposed technique can be instrumental in achieving sparsity-related energy/latency benefits without compromising on fault tolerance.

As language models (LMs) scale, they develop many novel behaviors, good and bad, exacerbating the need to evaluate how they behave. Prior work creates evaluations with crowdwork (which is time-consuming and expensive) or existing data sources (which are not always available). Here, we automatically generate evaluations with LMs. We explore approaches with varying amounts of human effort, from instructing LMs to write yes/no questions to making complex Winogender schemas with multiple stages of LM-based generation and filtering. Crowdworkers rate the examples as highly relevant and agree with 90-100% of labels, sometimes more so than corresponding human-written datasets. We generate 154 datasets and discover new cases of inverse scaling where LMs get worse with size. Larger LMs repeat back a dialog user's preferred answer ("sycophancy") and express greater desire to pursue concerning goals like resource acquisition and goal preservation. We also find some of the first examples of inverse scaling in RL from Human Feedback (RLHF), where more RLHF makes LMs worse. For example, RLHF makes LMs express stronger political views (on gun rights and immigration) and a greater desire to avoid shut down. Overall, LM-written evaluations are high-quality and let us quickly discover many novel LM behaviors.

Active target sensing is the task of discovering and classifying an unknown number of targets in an environment and is critical in search-and-rescue missions. This paper develops a deep reinforcement learning approach to plan informative trajectories that increase the likelihood for an uncrewed aerial vehicle (UAV) to discover missing targets. Our approach efficiently (1) explores the environment to discover new targets, (2) exploits its current belief of the target states and incorporates inaccurate sensor models for high-fidelity classification, and (3) generates dynamically feasible trajectories for an agile UAV by employing a motion primitive library. Extensive simulations on randomly generated environments show that our approach is more efficient in discovering and classifying targets than several other baselines. A unique characteristic of our approach, in contrast to heuristic informative path planning approaches, is that it is robust to varying amounts of deviations of the prior belief from the true target distribution, thereby alleviating the challenge of designing heuristics specific to the application conditions.

As AI systems become more capable, we would like to enlist their help to supervise other AIs. We experiment with methods for training a harmless AI assistant through self-improvement, without any human labels identifying harmful outputs. The only human oversight is provided through a list of rules or principles, and so we refer to the method as 'Constitutional AI'. The process involves both a supervised learning and a reinforcement learning phase. In the supervised phase we sample from an initial model, then generate self-critiques and revisions, and then finetune the original model on revised responses. In the RL phase, we sample from the finetuned model, use a model to evaluate which of the two samples is better, and then train a preference model from this dataset of AI preferences. We then train with RL using the preference model as the reward signal, i.e. we use 'RL from AI Feedback' (RLAIF). As a result we are able to train a harmless but non-evasive AI assistant that engages with harmful queries by explaining its objections to them. Both the SL and RL methods can leverage chain-of-thought style reasoning to improve the human-judged performance and transparency of AI decision making. These methods make it possible to control AI behavior more precisely and with far fewer human labels.

We present a differentiable formulation of rigid-body contact dynamics for objects and robots represented as compositions of convex primitives. Existing optimization-based approaches simulating contact between convex primitives rely on a bilevel formulation that separates collision detection and contact simulation. These approaches are unreliable in realistic contact simulation scenarios because isolating the collision detection problem introduces contact location non-uniqueness. Our approach combines contact simulation and collision detection into a unified single-level optimization problem. This disambiguates the collision detection problem in a physics-informed manner. Compared to previous differentiable simulation approaches, our formulation features improved simulation robustness and a reduction in computational complexity by more than an order of magnitude. We illustrate the contact and collision differentiability on a robotic manipulation task requiring optimization-through-contact. We provide a numerically efficient implementation of our formulation in the Julia language called Silico.jl.

Edge computing is becoming more and more popular among researchers who seek to take advantage of the edge resources and the minimal time delays, in order to run their robotic applications more efficiently. Recently, many edge architectures have been proposed, each of them having their advantages and disadvantages, depending on each application. In this work, we present two different edge architectures for controlling the trajectory of an Unmanned Aerial Vehicle (UAV). The first architecture is based on docker containers and the second one is based on kubernetes, while the main framework for operating the robot is the Robotic Operating System (ROS). The efficiency of the overall proposed scheme is being evaluated through extended simulations for comparing the two architectures and the overall results obtained.

Many practical applications, such as recommender systems and learning to rank, involve solving multiple similar tasks. One example is learning of recommendation policies for users with similar movie preferences, where the users may still rank the individual movies slightly differently. Such tasks can be organized in a hierarchy, where similar tasks are related through a shared structure. In this work, we formulate this problem as a contextual off-policy optimization in a hierarchical graphical model from logged bandit feedback. To solve the problem, we propose a hierarchical off-policy optimization algorithm (HierOPO), which estimates the parameters of the hierarchical model and then acts pessimistically with respect to them. We instantiate HierOPO in linear Gaussian models, for which we also provide an efficient implementation and analysis. We prove per-task bounds on the suboptimality of the learned policies, which show a clear improvement over not using the hierarchical model. We also evaluate the policies empirically. Our theoretical and empirical results show a clear advantage of using the hierarchy over solving each task independently.

Parameter-efficient methods (like Prompt or Adapters) for adapting pre-trained language models to downstream tasks have been popular recently. However, hindrances still prevent these methods from reaching their full potential. For example, two significant challenges are few-shot adaptation and cross-task generalization ability. To tackle these issues, we propose a general framework to enhance the few-shot adaptation and cross-domain generalization ability of parameter-efficient methods. In our framework, we prime the self-supervised model for parameter-efficient methods to rapidly adapt to various downstream few-shot tasks. To evaluate the authentic generalization ability of these parameter-efficient methods, we conduct experiments on a few-shot cross-domain benchmark containing 160 diverse NLP tasks. The experiment result reveals that priming by tuning PLM only with extra training tasks leads to the best performance. Also, we perform a comprehensive analysis of various parameter-efficient methods under few-shot cross-domain scenarios.

One of the recent advances in surgical AI is the recognition of surgical activities as triplets of (instrument, verb, target). Albeit providing detailed information for computer-assisted intervention, current triplet recognition approaches rely only on single frame features. Exploiting the temporal cues from earlier frames would improve the recognition of surgical action triplets from videos. In this paper, we propose Rendezvous in Time (RiT) - a deep learning model that extends the state-of-the-art model, Rendezvous, with temporal modeling. Focusing more on the verbs, our RiT explores the connectedness of current and past frames to learn temporal attention-based features for enhanced triplet recognition. We validate our proposal on the challenging surgical triplet dataset, CholecT45, demonstrating an improved recognition of the verb and triplet along with other interactions involving the verb such as (instrument, verb). Qualitative results show that the RiT produces smoother predictions for most triplet instances than the state-of-the-arts. We present a novel attention-based approach that leverages the temporal fusion of video frames to model the evolution of surgical actions and exploit their benefits for surgical triplet recognition.