This interactive paper aims to provide an intuitive understanding of the self-calibrating interface paradigm. Under this paradigm, you can choose how to use an interface which can adapt to your preferences on the fly. We introduce a PIN entering task and gradually release constraints, moving from a pre-calibrated interface to a self-calibrating interface while increasing the complexity of input modalities from buttons, to points on a map, to sketches, and finally to spoken words. This is not a traditional research paper with a hypothesis and experimental results to support claims; the research supporting this work has already been done and we refer to it extensively in the later sections. Instead, our aim is to walk you through an intriguing interaction paradigm in small logical steps with supporting illustrations, interactive demonstrations, and videos to reinforce your learning. We designed this paper for the enjoyments of curious minds of any backgrounds, it is written in plain English and no prior knowledge is necessary. All demos are available online at openvault.jgrizou.com and linked individually in the paper.
相關內容
IFIP TC13 Conference on Human-Computer Interaction是人機交互領域的研究者和實踐者展示其工作的重要平臺。多年來，這些會議吸引了來自幾個國家和文化的研究人員。官網鏈接： ·
泛化理論
·
Extensibility
·
INFORMS
·
Analysis
·
Recent progress in information retrieval finds that embedding query and document representation into multi-vector yields a robust bi-encoder retriever on out-of-distribution datasets. In this paper, we explore whether late interaction, the simplest form of multi-vector, is also helpful to neural rerankers that only use the [CLS] vector to compute the similarity score. Although intuitively, the attention mechanism of rerankers at the previous layers already gathers the token-level information, we find adding late interaction still brings an extra 5% improvement in average on out-of-distribution datasets, with little increase in latency and no degradation in in-domain effectiveness. Through extensive experiments and analysis, we show that the finding is consistent across different model sizes and first-stage retrievers of diverse natures and that the improvement is more prominent on longer queries.
Search systems on the Web rely on user input to generate relevant results. Since early information retrieval systems, users are trained to issue keyword searches and adapt to the language of the system. Recent research has shown that users often withhold detailed information about their initial information need, although they are able to express it in natural language. We therefore conduct a user study (N = 139) to investigate how four different design variants of search interfaces can encourage the user to reveal more information. Our results show that a chatbot-inspired search interface can increase the number of mentioned product attributes by 84% and promote natural language formulations by 139% in comparison to a standard search bar interface.
We study a foundational variant of Valiant and Vapnik and Chervonenkis' Probably Approximately Correct (PAC)-Learning in which the adversary is restricted to a known family of marginal distributions $\mathscr{P}$. In particular, we study how the PAC-learnability of a triple $(\mathscr{P},X,H)$ relates to the learners ability to infer \emph{distributional} information about the adversary's choice of $D \in \mathscr{P}$. To this end, we introduce the `unsupervised' notion of \emph{TV-Learning}, which, given a class $(\mathscr{P},X,H)$, asks the learner to approximate $D$ from unlabeled samples with respect to a natural class-conditional total variation metric. In the classical distribution-free setting, we show that TV-learning is \emph{equivalent} to PAC-Learning: in other words, any learner must infer near-maximal information about $D$. On the other hand, we show this characterization breaks down for general $\mathscr{P}$, where PAC-Learning is strictly sandwiched between two approximate variants we call `Strong' and `Weak' TV-learning, roughly corresponding to unsupervised learners that estimate most relevant distances in $D$ with respect to $H$, but differ in whether the learner \emph{knows} the set of well-estimated events. Finally, we observe that TV-learning is in fact equivalent to the classical notion of \emph{uniform estimation}, and thereby give a strong refutation of the uniform convergence paradigm in supervised learning.
Many tools empower analysts and data scientists to consume analysis results in a visual interface, such as a dashboard. When the underlying data changes, these results need to be updated, but this update can take a long time -- all while the user continues to explore the results. In this context, tools can either (i) hide away results that haven't been updated, hindering exploration; (ii) make the updated results immediately available to the user (on the same screen as old results), leading to confusion and incorrect insights; or (iii) present old -- and therefore stale -- results to the user during the update. To help users reason about these options and others, and make appropriate trade-offs, we introduce Transactional Panorama, a formal framework that adopts transactions to jointly model the system refreshing the analysis results and the user interacting with them. We introduce three key properties that are important for user perception in this context, visibility (allowing users to continuously explore results), consistency (ensuring that results resented are from the same version of the data), and monotonicity (making sure that results don't "go back in time"). Within transactional panorama, we characterize all of the feasible property combinations, design new mechanisms (that we call lenses) for presenting analysis results to the user while preserving a given property combination, formally prove their relative orderings for various performance criteria and discuss their use cases. We propose novel algorithms to preserve each property combination and efficiently present fresh analysis results. We implement our transactional panorama framework in a popular, open-source BI tool, illustrate the relative performance implications of different lenses, demonstrate the benefits of the novel lenses, and outline the performance improvement by our optimizations.
On end-to-end driving, a large amount of expert driving demonstrations is used to train an agent that mimics the expert by predicting its control actions. This process is self-supervised on vehicle signals (e.g., steering angle, acceleration) and does not require extra costly supervision (human labeling). Yet, the improvement of existing self-supervised end-to-end driving models has mostly given room to modular end-to-end models where labeling data intensive format such as semantic segmentation are required during training time. However, we argue that the latest self-supervised end-to-end models were developed in sub-optimal conditions with low-resolution images and no attention mechanisms. Further, those models are confined with limited field of view and far from the human visual cognition which can quickly attend far-apart scene features, a trait that provides an useful inductive bias. In this context, we present a new end-to-end model, trained by self-supervised imitation learning, leveraging a large field of view and a self-attention mechanism. These settings are more contributing to the agent's understanding of the driving scene, which brings a better imitation of human drivers. With only self-supervised training data, our model yields almost expert performance in CARLA's Nocrash metrics and could be rival to the SOTA models requiring large amounts of human labeled data. To facilitate further research, our code will be released.
Attackers may attempt exploiting Internet of Things (IoT) devices to operate them unduly as well as to gather personal data of the legitimate device owners'. Vulnerability Assessment and Penetration Testing (VAPT) sessions help to verify the effectiveness of the adopted security measures. However, VAPT over IoT devices, namely VAPT targeted at IoT devices, is an open research challenge due to the variety of target technologies and to the creativity it may require. Therefore, this article aims at guiding penetration testers to conduct VAPT sessions over IoT devices by means of a new cyber Kill Chain (KC) termed PETIoT. Several practical applications of PETIoT confirm that it is general, while its main novelty lies in the combination of attack and defence steps. PETIoT is demonstrated on a relevant example, the best-selling IP camera on Amazon Italy, the TAPO C200 by TP-Link, assuming an attacker who sits on the same network as the device's in order to assess all the network interfaces of the device. Additional knowledge is generated in terms of three zero-day vulnerabilities found and practically exploited on the camera, one of these with High severity and the other two with Medium severity by the CVSS standard. These are camera Denial of Service (DoS), motion detection breach and video stream breach. The application of PETIoT culminates with the proof-of-concept of a home-made fix, based on an inexpensive Raspberry Pi 4 Model B device, for the last vulnerability. Ultimately, our responsible disclosure with the camera vendor led to the release of a firmware update that fixes all found vulnerabilities, confirming that PetIoT has valid impact in real-world scenarios.
Advances in artificial intelligence often stem from the development of new environments that abstract real-world situations into a form where research can be done conveniently. This paper contributes such an environment based on ideas inspired by elementary Microeconomics. Agents learn to produce resources in a spatially complex world, trade them with one another, and consume those that they prefer. We show that the emergent production, consumption, and pricing behaviors respond to environmental conditions in the directions predicted by supply and demand shifts in Microeconomics. We also demonstrate settings where the agents' emergent prices for goods vary over space, reflecting the local abundance of goods. After the price disparities emerge, some agents then discover a niche of transporting goods between regions with different prevailing prices -- a profitable strategy because they can buy goods where they are cheap and sell them where they are expensive. Finally, in a series of ablation experiments, we investigate how choices in the environmental rewards, bartering actions, agent architecture, and ability to consume tradable goods can either aid or inhibit the emergence of this economic behavior. This work is part of the environment development branch of a research program that aims to build human-like artificial general intelligence through multi-agent interactions in simulated societies. By exploring which environment features are needed for the basic phenomena of elementary microeconomics to emerge automatically from learning, we arrive at an environment that differs from those studied in prior multi-agent reinforcement learning work along several dimensions. For example, the model incorporates heterogeneous tastes and physical abilities, and agents negotiate with one another as a grounded form of communication.
The dominating NLP paradigm of training a strong neural predictor to perform one task on a specific dataset has led to state-of-the-art performance in a variety of applications (eg. sentiment classification, span-prediction based question answering or machine translation). However, it builds upon the assumption that the data distribution is stationary, ie. that the data is sampled from a fixed distribution both at training and test time. This way of training is inconsistent with how we as humans are able to learn from and operate within a constantly changing stream of information. Moreover, it is ill-adapted to real-world use cases where the data distribution is expected to shift over the course of a model's lifetime. The first goal of this thesis is to characterize the different forms this shift can take in the context of natural language processing, and propose benchmarks and evaluation metrics to measure its effect on current deep learning architectures. We then proceed to take steps to mitigate the effect of distributional shift on NLP models. To this end, we develop methods based on parametric reformulations of the distributionally robust optimization framework. Empirically, we demonstrate that these approaches yield more robust models as demonstrated on a selection of realistic problems. In the third and final part of this thesis, we explore ways of efficiently adapting existing models to new domains or tasks. Our contribution to this topic takes inspiration from information geometry to derive a new gradient update rule which alleviate catastrophic forgetting issues during adaptation.
The Q-learning algorithm is known to be affected by the maximization bias, i.e. the systematic overestimation of action values, an important issue that has recently received renewed attention. Double Q-learning has been proposed as an efficient algorithm to mitigate this bias. However, this comes at the price of an underestimation of action values, in addition to increased memory requirements and a slower convergence. In this paper, we introduce a new way to address the maximization bias in the form of a "self-correcting algorithm" for approximating the maximum of an expected value. Our method balances the overestimation of the single estimator used in conventional Q-learning and the underestimation of the double estimator used in Double Q-learning. Applying this strategy to Q-learning results in Self-correcting Q-learning. We show theoretically that this new algorithm enjoys the same convergence guarantees as Q-learning while being more accurate. Empirically, it performs better than Double Q-learning in domains with rewards of high variance, and it even attains faster convergence than Q-learning in domains with rewards of zero or low variance. These advantages transfer to a Deep Q Network implementation that we call Self-correcting DQN and which outperforms regular DQN and Double DQN on several tasks in the Atari 2600 domain.
In this paper, we present an accurate and scalable approach to the face clustering task. We aim at grouping a set of faces by their potential identities. We formulate this task as a link prediction problem: a link exists between two faces if they are of the same identity. The key idea is that we find the local context in the feature space around an instance (face) contains rich information about the linkage relationship between this instance and its neighbors. By constructing sub-graphs around each instance as input data, which depict the local context, we utilize the graph convolution network (GCN) to perform reasoning and infer the likelihood of linkage between pairs in the sub-graphs. Experiments show that our method is more robust to the complex distribution of faces than conventional methods, yielding favorably comparable results to state-of-the-art methods on standard face clustering benchmarks, and is scalable to large datasets. Furthermore, we show that the proposed method does not need the number of clusters as prior, is aware of noises and outliers, and can be extended to a multi-view version for more accurate clustering accuracy.
小貼士
登錄享
相關主題
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191