AAAI 2022 Papers

Due to the surge of cloud-assisted AI services, the problem of designing resilient prediction serving systems that can effectively cope with stragglers and minimize response delays has attracted much interest. The common approach for tackling this problem is replication which assigns the same prediction task to multiple workers. This approach, however, is inefficient and incurs significant resource overheads. Hence, a learning-based approach known as parity model (ParM) has been recently proposed which learns models that can generate ``parities’’ for a group of predictions to reconstruct the predictions of the slow/failed workers. While this learning-based approach is more resource-efficient than replication, it is tailored to the specific model hosted by the cloud and is particularly suitable for a small number of queries (typically less than four) and tolerating very few stragglers (mostly one). Moreover, ParM does not handle Byzantine adversarial workers. We propose a different approach, named Approximate Coded Inference (ApproxIFER), that does not require training any parity models, hence it is agnostic to the model hosted by the cloud and can be readily applied to different data domains and model architectures. Compared with earlier works, ApproxIFER can handle a general number of stragglers and scales significantly better with the number of queries. Furthermore, ApproxIFER is robust against Byzantine workers. Our extensive experiments on a large number of datasets and model architectures show significant degraded mode accuracy improvement by up to 58% over ParM.

Answer Set Programming (ASP) is a framework in artificial intelligence and knowledge representation for declarative modeling and problem solving. Modern ASP solvers focus on the computation or enumeration of answer sets. However, a variety of probabilistic applications in reasoning or logic programming require counting answer sets. While counting can be done by enumeration, simple enumeration becomes immediately infeasible if the number of solutions is high. On the other hand, approaches to exact counting are of high worst-case complexity. In fact, in propositional model counting, exact counting becomes impractical. In this work, we present a scalable approach to approximate counting for answer set programming. Our approach is based on systematically adding XOR constraints to ASP programs, which divide the search space. We prove that adding random XOR constraints partitions the answer sets of an ASP program. In practice, we use a Gaussian elimination-based approach by lifting ideas from SAT to ASP and integrating it into a state of the art ASP solver, which we call ApproxASP. Finally, our experimental evaluation shows the scalability of our approach over the existing ASP systems.

We investigate approval-based committee voting with incomplete information about the approval preferences of voters. We consider several models of incompleteness where each voter partitions the set of candidates into approved, disapproved, and unknown candidates, possibly with ordinal preference constraints among candidates in the latter category. This captures scenarios where voters have not evaluated all candidates and/or it is unknown where voters draw the threshold between approved and disapproved candidates. We study the complexity of some fundamental computational problems for a number of classic approval-based committee voting rules including Proportional Approval Voting and Chamberlin-Courant. These problems include that of determining whether a given set of candidates is a possible or necessary winning committee and whether it forms a committee that possibly or necessarily satisfies representation axioms. We also consider the problem whether a given candidate is possibly or necessarily a member of the winning committee.

Multi-Agent Path Finding (MAPF) is the problem of finding a set of collision-free paths for a team of agents in a common environment. MAPF is NP-hard to solve optimally and, in some cases, also bounded-suboptimally. It is thus time-consuming for (bounded-sub)optimal solvers to solve large MAPF instances. Anytime algorithms find solutions quickly for large instances and then improve them to close-to-optimal ones over time. In this paper, we improve the current state-of-the-art anytime solver MAPF-LNS, that first finds an initial solution fast and then repeatedly replans the paths of subsets of agents via Large Neighborhood Search (LNS). It generates the subsets of agents for replanning by randomized destroy heuristics, but not all of them increase the solution quality substantially. We propose to use machine learning to learn how to select a subset of agents from a collection of subsets, such that replanning increases the solution quality more. We show experimentally that our solver, MAPF-ML-LNS, significantly outperforms MAPF-LNS on the standard MAPF benchmark set in terms of both the speed of improving the solution and the final solution quality.

Under data distributions which may be heavy-tailed, many stochastic gradient-based learning algorithms are driven by feedback queried at points with almost no performance guarantees on their own. Here we explore a modified "anytime online-to-batch" mechanism which for smooth objectives admits high-probability error bounds while requiring only lower-order moment bounds on the stochastic gradients. Using this conversion, we can derive a wide variety of "anytime robust" procedures, for which the task of performance analysis can be effectively reduced to regret control, meaning that existing regret bounds (for the bounded gradient case) can be robustified and leveraged in a straightforward manner. As a direct takeaway, we obtain an easily implemented stochastic gradient-based algorithm for which all queried points formally enjoy sub-Gaussian error bounds, and in practice show noteworthy gains on real-world data applications.

Ontology-based query answering with existential rules is well understood and implemented for positive queries, in particular conjunctive queries. For queries with negation, however, there is no agreed-upon semantics or standard implementation. This problem is unknown for simpler rule languages, such as Datalog, where it is intuitive and practical to evaluate negative queries over the least model. This fails for existential rules, which instead of a single least model have multiple universal models that may not lead to the same results for negative queries. We therefore propose universal core models as a basis for a meaningful (non-monotonic) semantics for queries with negation. Since cores are hard to compute, we identify syntactic conditions (on rules and queries) under which our core-based semantics can equivalently be obtained for other universal models, such as those produced by practical chase algorithms. Finally, we use our findings to propose a semantics for a broad class of existential rules with negation.

Recent neural rendering techniques have greatly benefited image-based modeling and relighting tasks. They provide a continuous, compact, and parallelable representation by modeling the plenoptic function as multilayer perceptrons (MLPs). However, vanilla MLPs suffer from spectral biases on multidimensional datasets. Recent rescues based on isotropic Fourier features mapping mitigate the problem but still fall short of handling heterogeneity across different dimensions, causing imbalanced regression and visual artifacts such as excessive blurs. We present an anisotropic random Fourier features (RFF) mapping scheme to tackle spectral biases. We first analyze the influence of bandwidth from a different perspective: we show that the optimal bandwidth exhibits strong correlations with the frequency spectrum of the training data across various dimensions. We then introduce an anisotropic feature mapping scheme with multiple bandwidths to model the multidimensional signal characteristics. We further propose an efficient bandwidth searching scheme through iterative golden-section search that can significantly reduce the training overload from polynomial time to logarithm. Our anisotropic scheme directly applies to neural surface light-field rendering and image-based relighting. Comprehensive experiments show that our scheme can more faithfully model lighting conditions and object features as well as preserve fine texture details and smooth view transitions even when angular and spatial samples are highly imbalanced.

Maximum Inner Product Search (MIPS) plays an important role in many applications ranging from information retrieval, recommender systems to natural language processing and machine learning. However, exhaustive MIPS is often expensive and impractical when there are a large number of candidate items. The state-of-the-art approximated MIPS is product quantization with a score-aware loss, which weighs more heavily on items with larger inner product scores. However, it is challenging to extend the score-aware loss for additive quantization due to parallel-orthogonal decomposition of residual error. Learning additive quantization with respect to this loss is important since additive quantization can achieve a lower approximation error than product quantization. To this end, we propose a quantization method called Anisotropic Additive Quantization to combine the score-aware anisotropic loss and additive quantization. To efficiently update the codebooks in this algorithm, we develop a new alternating optimization algorithm. The proposed algorithm is extensively evaluated on three real-world datasets. The experimental results show that it outperforms the state-of-the-art baselines with respect to approximate search accuracy while guaranteeing a similar retrieval efficiency.

Within the field of face recognition (FR), it is widely accepted that the key objective is to optimize the entire feature space in the training process and acquire robust feature representations. However, most real-world FR systems tend to operate at a pre-defined False Accept Rate (FAR), and the corresponding True Accept Rate (TAR) represents the performance of the FR systems, which indicates that the optimization on the pre-defined FAR is more meaningful and important in the practical evaluation process. In this paper, we call the predefined FAR as Anchor FAR, and we argue that the existing FR loss functions cannot guarantee the optimal TAR under the Anchor FAR, which impedes further improvements of FR systems. To this end, we propose AnchorFace to bridge the aforementioned gap between the training and practical evaluation process for FR. Given the Anchor FAR, AnchorFace can boost the performance of FR systems by directly optimizing the non-differentiable FR evaluation metrics. Specifically, in AnchorFace, we first calculate the similarities of the positive and negative pairs based on both the features of the current batch and the stored features in the maintained online-updating set. Then, we generate the differentiable TAR loss and FAR loss using a soften strategy. Our AnchorFace can be readily integrated into most existing FR loss functions, and extensive experimental results on multiple benchmark datasets demonstrate the effectiveness of AnchorFace.

In this paper, we propose a novel query design for the transformer-based object detection. In previous transformer-based detectors, the object queries are a set of learned embeddings. However, each learned embedding does not have an explicit physical meaning and we cannot explain where it will focus on. It is difficult to optimize as the prediction slot of each object query does not have a specific mode. In other words, each object query will not focus on a specific region. To solve these problems, in our query design, object queries are based on anchor points, which are widely used in CNN-based detectors. So each object query focuses on the objects near the anchor point. Moreover, our query design can predict multiple objects at one position to solve the difficulty: ``one region, multiple objects''. In addition, we design an attention variant, which can reduce the memory cost while achieving similar or better performance than the standard attention in DETR. Thanks to the query design and the attention variant, the proposed detector that we called Anchor DETR, can achieve better performance and run faster than the DETR with 10x fewer training epochs. For example, it achieves 44.2 AP with 19 FPS on the MSCOCO dataset when using the ResNet50-DC5 feature for training 50 epochs. Extensive experiments on the MSCOCO benchmark prove the effectiveness of the proposed methods. Code is available at https://github.com/megvii-research/AnchorDETR.

MAX k-SAT is one of the archetypal NP-hard problems. Its variation called random MAX k-SAT problem was introduced in order to understand how hard it is to solve instances of the problem on average. The most common model to sample random instances is the uniform model, which has received a large amount of attention. However, the uniform model often fails to capture important structural properties we observe in the real-world instances. To address these limitations, a more general (in a certain sense) model has been proposed, the configuration model, which is able to produce instances with an arbitrary distribution of variables' degrees, and so can simulate biases in instances appearing in various applications. Our overall goal is to expand the theory built around the uniform model to the more general configuration model for a wide range of degree distributions. This includes locating satisfiability thresholds and analysing the performance of the standard heuristics applied to instances sampled from the configuration model. In this paper we analyse the performance of the pure literal elimination rule. We provide an equation that given an underlying degree distribution gives the number of clauses the pure literal elimination rule satisfies w.h.p. We also show how the distribution of variable degrees changes over time as the algorithm is being executed.

Despite significant improvements on the image generation performance of Generative Adversarial Networks (GANs), generations with low visual fidelity still have been observed. As widely used metrics for GANs focus more on the overall performance of the model, evaluation on the quality of individual generations or detection of defective generations is challenging. While recent studies try to detect featuremap units that cause artifacts and evaluate individual samples, these approaches require additional resources such as external networks or a number of training data to approximate the real data manifold. In this work, we propose the concept of local activation, and devise a metric on the local activation to detect artifact generations without additional supervision. We empirically verify that our approach can detect and correct artifact generations from GANs with various datasets. Finally, we discuss a geometrical analysis to partially reveal the relation between the proposed concept and low visual fidelity.

This paper proposes a new variant of multi-play MAB model, to capture important factors of the sequential user-centric selection problem arising from mobile edge computing, ridesharing applications, etc. In the proposed model, each arm is associated with discrete units of resources, each play is associate with movement costs and multiple plays can pull the same arm simultaneously. To learn the optimal action profile (an action profile prescribes the arm that each play pulls), there are two challenges: (1) the number of action profiles is large, i.e., M^K, where K and M denote the number of plays and arms respectively; (2) feedbacks on action profiles are not available, but instead feedbacks on some model parameters can be observed. To address the first challenge, we formulate a completed weighted bipartite graph to capture key factors of the offline decision problem with given model parameters. We identify the correspondence between action profiles and a special class of matchings of the graph. We also identify a dominance structure of this class of matchings. This correspondence and dominance structure enable us to design an algorithm named OffOptActPrf to locate the optimal action efficiently. To address the second challenge, we design an OnLinActPrf algorithm. We design estimators for model parameters and use these estimators to design a Quasi-UCB index for each action profile. The OnLinActPrf uses OffOptActPrf as a subroutine to select the action profile with the largest Quasi-UCB index. We conduct extensive experiments to validate the efficiency of OnLinActPrf.

In this work we consider the problem of regret minimization for logistic bandits. The main challenge of logistic bandits is reducing the dependence on a potentially large problem dependent constant that can at worst scale exponentially with the norm of the unknown parameter vector. Previous works have applied self-concordance of the logistic function to remove this worst-case dependence providing regret guarantees that move the reduce the dependence on this worst case parameter to lower order terms with only polylogarithmic dependence on the main term and as well as linear dependence on the dimension of the unknown parameter. This work improves upon the prior art by 1) removing all scaling of the worst case term on the main term and 2) reducing the dependence on the dependence to scale with the square root of dimension in the fixed arm setting by employing an experimental design procedure. Our regret bound in fact takes a tighter instance (i.e., gap) dependent regret bound for the first time in logistic bandits. We also propose a new warmup sampling algorithm that can dramatically reduce the lower order term in the regret in general and prove that it can exponentially reduce the lower order term's dependency on the worst case parameter in some instances. Finally, we discuss the impact of the bias of the MLE on the logistic bandit problem in d dimensions, providing an example where d^2 lower order regret (cf., it is d for linear bandits) may not be improved as long as the MLE is used and how bias-corrected estimators may be used to make it closer to d.

The Maximum k-Defective Clique Problem (MDCP), as a clique relaxation model, has been used to solve various problems. Because it is a hard computational task, previous works can hardly solve the MDCP for massive sparse graphs derived from real-world applications. In this work, we propose a novel branch-and-bound algorithm to solve the MDCP based on several new techniques. First, we propose two new upper bounds of the MDCP as well as corresponding reduction rules to remove redundant vertices and edges. The proposed reduction rules are particularly useful for massive graphs. Second, we present another new upper bound by counting missing edges between fixed vertices and an unfixed vertex for cutting branches. We perform extensive computational experiments to evaluate our algorithm. Experimental results show that our reduction rules are very effective for removing redundant vertices and edges so that graphs are reduced greatly. Also, our algorithm can solve benchmark instances efficiently, and it has significantly better performance than state-of-the-art algorithms.

Clustering algorithms are often evaluated using metrics which compare with ground-truth cluster assignments, such as Rand index and NMI. Algorithm performance may vary widely for different hyperparameters, however, and thus model selection based on optimal performance for these metrics is discordant with how these algorithms are applied in practice, where labels are unavailable and tuning is often more art than science. It is therefore desirable to compare clustering algorithms not only on their optimally tuned performance, but also some notion of how realistic it would be to obtain this performance in practice. We propose an evaluation of clustering methods capturing this ease-of-tuning by modeling the expected best clustering score under a given computation budget. To encourage the adoption of the proposed metric alongside classic clustering evaluations, we provide an extensible benchmarking framework. We perform an extensive empirical evaluation of our proposed metric on popular clustering algorithms over a large collection of datasets from different domains, and observe that our new metric leads to several noteworthy observations.

Knowledge-based visual question answering (VQA) involves answering questions that require external knowledge not present in the image. Existing methods first retrieve knowledge from external resources, then reason over the selected knowledge, the input image, and question for answer prediction. However, this two-step approach could lead to mismatches that potentially limit the VQA performance. For example, the retrieved knowledge might be noisy and irrelevant to the question, and the re-embedded knowledge features during reasoning might deviate from their original meanings in the knowledge base (KB). To address this challenge, we propose PICa, a simple yet effective method that Prompts GPT3 via the use of Image Captions, for knowledge-based VQA. Inspired by GPT-3’s power in knowledge retrieval and question answering, instead of using structured KBs as in previous work, we treat GPT-3 as an implicit and unstructured KB that can jointly acquire and process relevant knowledge. Specifically, we first convert the image into captions (or tags) that GPT-3 can understand, then adapt GPT-3 to solve the VQA task in a few-shot manner by just providing a few in-context VQA examples. We further boost performance by carefully investigating: (i) what text formats best describe the image content, and (ii) how in-context examples can be better selected and used. PICa unlocks the first use of GPT-3 for multimodal tasks. By using only 16 examples, PICa surpasses the supervised state of the art by an absolute +8.6 points on the OK-VQA dataset. We also benchmark PICa on VQAv2, where PICa also shows a decent few-shot performance.

Recently, deep reinforcement learning (RL) has proven its feasibility in solving combinatorial optimization problems (COPs). The learning-to-rank techniques have been studied in the field of information retrieval. While several COPs can be formulated as the prioritization of input items, as is common in the information retrieval, it has not been fully explored how the learning-to-rank techniques can be incorporated into deep RL for COPs. In this paper, we present the learning-to-rank distillation-based COP framework, where a high-performance ranking policy obtained by RL for a COP can be distilled into a non-iterative, simple model, thereby achieving a low-latency COP solver. Specifically, we employ the approximated ranking distillation to render a score-based ranking model learnable via gradient descent. Furthermore, we use the efficient sequence sampling to improve the inference performance with a limited delay. With the framework, we demonstrate that a distilled model not only achieves comparable performance to its respective, high-performance RL, but also provides several times faster inferences. We evaluate the framework with several COPs such as priority-based task scheduling and multidimensional knapsack, demonstrating the benefits of the framework in terms of inference latency and performance.

We study a model of preference revision in which a prior preference over a set of alternatives is adjusted in order to accommodate input from an authoritative source, while maintaining certain structural constraints (e.g., transitivity, completeness), and without giving up more information than strictly necessary. We analyze this model under two aspects: the first allows us to capture natural distance-based operators, at the cost of a mismatch between the input and output formats of the revision operator. Requiring the input and output to be aligned yields a second type of operator, which we characterize using preferences on the comparisons in the prior preference Prefence revision is set in a logic-based framework and using the formal machinery of belief change, along the lines of the well-known AGM approach: we propose rationality postulates for each of the two versions of our model and derive representation results, thus situating preference revision within the larger family of belief change operators.