Kilian Weinberger

• PH.D.
• Professor (Associate) at Cornell University

Diffusion models have achieved great success in modeling continuous data modalities such as images, audio, and video, but have seen limited use in discrete domains such as language. Recent attempts to adapt diffusion to language have presented diffusion as an alternative to autoregressive language generation. We instead view diffusion as a compleme...

The information content of crystalline materials becomes astronomical when collective electronic behavior and their fluctuations are taken into account. In the past decade, improvements in source brightness and detector technology at modern X-ray facilities have allowed a dramatically increased fraction of this information to be captured. Now, the...

The information content of crystalline materials becomes astronomical when distortions, defects, phase heterogeneity, and collective electronic behavior are taken into account. In the past decade, improvements in source brightness and detector technology at modern x-ray facilities have allowed a dramatically increased fraction of this information t...

A widely deployed method for reducing the training time of deep neural networks is to normalize activations at each layer. Although various normalization schemes have been proposed, they all follow a common theme: normalize across spatial dimensions and discard the extracted statistics. In this paper, we propose a novel normalization method that no...

Detecting objects such as cars and pedestrians in 3D plays an indispensable role in autonomous driving. Existing approaches largely rely on expensive LiDAR sensors for accurate depth information. While recently pseudo-LiDAR has been introduced as a promising alternative, at a much lower cost based solely on stereo images, there is still a notable p...

Object segmentation in 3D point clouds is a critical task for robots capable of 3D perception. Despite the impressive performance of deep learning-based approaches on object segmentation in 2D images, deep learning has not been applied nearly as successfully for 3D point cloud segmentation. Deep networks generally require large amounts of labeled t...

Recent work has shown that convolutional networks can be substantially deeper, more accurate, and efficient to train if they contain shorter connections between layers close to the input and those close to the output. In this paper, we embrace this observation and introduce the Dense Convolutional Network (DenseNet), which connects each layer to ev...

We propose an intriguingly simple method for the construction of adversarial images in the black-box setting. In constrast to the white-box scenario, constructing black-box adversarial images has the additional constraint on query budget, and efficient attacks remain an open problem to date. With only the mild assumption of continuous-valued confid...

We propose BERTScore, an automatic evaluation metric for text generation. Analogous to common metrics, \method computes a similarity score for each token in the candidate sentence with each token in the reference. However, instead of looking for exact matches, we compute similarity using contextualized BERT embeddings. We evaluate on several machin...

Recent studies have discovered the vulnerability of Deep Neural Networks (DNNs) to adversarial examples, which are imperceptible to humans but can easily fool DNNs. Existing methods for crafting adversarial examples are mainly based on adding small-magnitude perturbations to the original images so that the generated adversarial examples are constra...

Gaussian processes (GPs) are flexible models with state-of-the-art performance on many impactful applications. However, computational constraints with standard inference procedures have limited exact GPs to problems with fewer than about ten thousand training points, necessitating approximations for larger datasets. In this paper, we develop a scal...

In this technical report, we introduce FastFusionNet, an efficient variant of FusionNet [12]. FusionNet is a high performing reading comprehension architecture, which was designed primarily for maximum retrieval accuracy with less regard towards computational requirements. For FastFusionNets we remove the expensive CoVe layers [21] and substitute t...

Graph Convolutional Networks (GCNs) and their variants have experienced significant attention and have become the de facto methods for learning graph representations. GCNs derive inspiration primarily from recent deep learning approaches, and as a result, may inherit unnecessary complexity and redundant computation. In this paper, we reduce this ex...

As machine learning transitions increasingly towards real world applications controlling the test-time cost of algorithms becomes more and more crucial. Recent work, such as the Greedy Miser and Speedboost, incorporate test-time budget constraints into the training procedure and learn classifiers that provably stay within budget (in expectation). H...

A feature selection algorithm should ideally satisfy four conditions: reliably extract relevant features; be able to identify non-linear feature interactions; scale linearly with the number of features and dimensions; allow the incorporation of known sparsity structure. In this work we propose a novel feature selection algorithm, Gradient Boosted F...

3D object detection is an essential task in autonomous driving. Recent techniques excel with highly accurate detection rates, provided the 3D input data is obtained from precise but expensive LiDAR technology. Approaches based on cheaper monocular or stereo imagery data have, until now, resulted in drastically lower accuracies --- a gap that is com...

Many real-world applications of stereo depth estimation in robotics require the generation of accurate disparity maps in real time under significant computational constraints. Current state-of-the-art algorithms can either generate accurate but slow mappings, or fast but inaccurate ones, and typically require far too many parameters for power- or m...

We present a data association method for vision-based multiple pedestrian tracking, using deep convolutional features to distinguish between different people based on their appearances. These re-identification (re-ID) features are learned such that they are invariant to transformations such as rotation, translation, and changes in the background, a...

Despite advances in scalable models, the inference tools used for Gaussian processes (GPs) have yet to fully capitalize on recent developments in machine learning hardware. We present an efficient and general approach to GP inference based on Blackbox Matrix-Matrix multiplication (BBMM). BBMM inference uses a modified batched version of the conjuga...

Recently, machine learning security has received significant attention. Many computer vision and speech recognition systems have been compromised by adversarially but imperceptibly perturbed input. To identify potential perturbations, attackers search the high dimensional input space to find directions in which the model lacks robustness. The expon...

Evaluating generative adversarial networks (GANs) is inherently challenging. In this paper, we revisit several representative sample-based evaluation metrics for GANs, and address the problem of how to evaluate the evaluation metrics. We start with a few necessary conditions for metrics to produce meaningful scores, such as distinguishing real from...

Not all people are equally easy to identify: color statistics might be enough for some cases while others might require careful reasoning about high- and low-level details. However, prevailing person re-identification(re-ID) methods use one-size-fits-all high-level embeddings from deep convolutional networks for all cases. This might limit their ac...

One of the most compelling features of Gaussian process (GP) regression is its ability to provide well calibrated posterior distributions. Recent advances in inducing point methods have drastically sped up marginal likelihood and posterior mean computations, leaving posterior covariance estimation and sampling as the remaining computational bottlen...

Recent work shows that inference for Gaussian processes can be performed efficiently using iterative methods that rely only on matrix-vector multiplications (MVMs). Structured Kernel Interpolation (SKI) exploits these techniques by deriving approximate kernels with very fast MVMs. Unfortunately, such strategies suffer badly from the curse of dimens...

Deep neural networks are increasingly used on mobile devices, where computational resources are limited. In this paper we develop CondenseNet, a novel network architecture with unprecedented efficiency. It combines dense connectivity between layers with a mechanism to remove unused connections. The dense connectivity facilitates feature re-use in t...

State-of-the-art deep reading comprehension models are dominated by recurrent neural nets. Their sequential nature is a natural fit for language, but it also precludes parallelization within an instances and often becomes the bottleneck for deploying such models to latency critical scenarios. This is particularly problematic for longer texts. Here...

The machine learning community has become increasingly concerned with the potential for bias and discrimination in predictive models, and this has motivated a growing line of work on what it means for a classification procedure to be "fair." In particular, we investigate the tension between minimizing error disparity across different population gro...

Recent work has shown that convolutional networks can be substantially deeper, more accurate, and efficient to train if they contain shorter connections between layers close to the input and those close to the output. In this paper, we embrace this observation and introduce the Dense Convolutional Network (DenseNet), which connects each layer to ev...

The DenseNet architecture is highly computationally efficient as a result of feature reuse. However, a naive DenseNet implementation can require a significant amount of GPU memory: If not properly managed, pre-activation batch normalization and contiguous convolution operations can produce feature maps that grow quadratically with network depth. In...

Confidence calibration -- the problem of predicting probability estimates representative of the true correctness likelihood -- is important for classification models in many applications. We discover that modern neural networks, unlike those from a decade ago, are poorly calibrated. Through extensive experiments, we observe that depth, width, weigh...

Ensembles of neural networks are known to be much more robust and accurate than individual networks. However, training multiple deep networks for model averaging is computationally expensive. In this paper, we propose a method to obtain the seemingly contradictory goal of ensembling multiple neural networks at no additional training cost. We achiev...

This paper studies convolutional networks that require limited computational resources at test time. We develop a new network architecture that performs on par with state-of-the-art convolutional networks, whilst facilitating prediction in two settings: (1) an anytime-prediction setting in which the network's prediction for one example is progressi...

We propose Deep Feature Interpolation (DFI), a new data-driven baseline for automatic high-resolution image transformation. As the name suggests, it relies only on simple linear interpolation of deep convolutional features from pre-trained convnets. We show that despite its simplicity, DFI can perform high-level semantic transformations like "make...

Very deep convolutional networks with hundreds of layers have led to significant reductions in error on competitive benchmarks. Although the unmatched expressiveness of the many layers can be highly desirable at test time, training very deep networks comes with its own set of challenges. The gradients can vanish, the forward flow often diminishes,...

Convolutional neural networks (CNN) are increasingly used in many areas of computer vision. They are particularly attractive because of their ability to "absorb" great quantities of labeled data through millions of parameters. However, as model sizes increase, so do the storage and memory requirements of the classifiers, hindering many applications...

Recent work has shown that convolutional networks can be substantially deeper, more accurate and efficient to train if they contain shorter connections between layers close to the input and those close to the output. In this paper we embrace this observation and introduce the Dense Convolutional Network (DenseNet), where each layer is directly conn...

Significance Since the early days of neuroscience, the relative merit of structural vs. functional network accounts in explaining neurological deficits has been intensely debated. Using a large stroke cohort and a machine-learning approach, we show that visual memory and verbal memory deficits are better predicted by functional connectivity than by...

In recent years deep neural networks have achieved great success in sentiment classification for English, thanks in part to the availability of copious annotated resources. Unfortunately, most other languages do not enjoy such an abundance of annotated data for sentiment analysis. To combat this problem, we propose the Adversarial Deep Averaging Ne...

In recent years great success has been achieved in sentiment classification for English, thanks in part to the availability of copious annotated resources. Unfortunately, most languages do not enjoy such an abundance of labeled data. To tackle the sentiment classification problem in low-resource languages without adequate annotated data, we propose...

Very deep convolutional networks with hundreds or more layers have lead to significant reductions in error on competitive benchmarks like the ImageNet or COCO tasks. Although the unmatched expressiveness of the many deep layers can be highly desirable at test time, training very deep networks comes with its own set of challenges. The gradients can...

Causal inference deals with identifying which random variables "cause" or control other random variables. Recent advances on the topic of causal inference based on tools from statistical estimation and machine learning have resulted in practical algorithms for causal inference. Causal inference has the potential to have significant impact on medica...

Stacked denoising autoencoders (SDAs) have been successfully used to learn new representations for domain adaptation. They have attained record accuracy on standard benchmark tasks of sentiment analysis across different text domains. SDAs learn robust data representations by reconstruction, recovering original features from data that are artificial...

Machine learning is increasingly used in high impact applications such as prediction of hospital re-admission, cancer screening or bio-medical research applications. As predictions become increasingly accurate, practitioners may be interested in identifying actionable changes to inputs in order to alter their class membership. For example, a doctor...

Previous studies suggest stable and robust control of a brain-computer interface (BCI) can be achieved using electrocorticography (ECoG). Translation of this technology from the laboratory to the real world requires additional methods that allow users operate their ECoG-based BCI autonomously. In such an environment, users must be able to perform a...

Pure-tone audiometry has been a staple of hearing assessments for decades. Many different procedures have been proposed for measuring thresholds with pure tones by systematically manipulating intensity one frequency at a time until a discrete threshold function is determined. The authors have developed a novel nonparametric approach for estimating...

Convolutional neural networks (CNN) are increasingly used in many areas of computer vision. They are particularly attractive because of their ability to "absorb" great quantities of labeled data through millions of parameters. However, as model sizes increase, so do the storage and memory requirements of the classifiers. We present a novel network...

As deep nets are increasingly used in applications suited for mobile devices, a fundamental dilemma becomes apparent: the trend in deep learning is to grow models to absorb ever-increasing data set sizes; however mobile devices are designed with very little memory and cannot store such large models. We present a novel network architecture, HashedNe...

Support vector machines (SVM) can classify data sets along highly non-linear decision boundaries because of the kernel-trick. This expressiveness comes at a price: During test-time, the SVM classifier needs to compute the kernel inner-product between a test sample and all support vectors. With large training data sets, the time required for this co...

Bayesian optimization has been shown to be a powerful tool for fine-tuning the hyper-parameters of a wide-variety of machine learning models. At the same time, the success of the field of machine learning has led practitioners in a wide variety of real-world settings to learn classifiers for practical problems. As real-world machine learning become...

Psychophysical detection tests are ubiquitous in the study of human sensation and the diagnosis and treatment of virtually all sensory impairments. In many of these settings, the goal is to recover, from a series of binary observations from a human subject, the latent function that describes the discriminability of a sensory stimulus over some rele...

Covariance matrices are an effective way to capture global spread across local interest points in images. Often, these image descriptors are more compact, robust and informative than, for example, bags of visual words. However, they are symmetric and positive definite (SPD) and therefore live on a non-Euclidean Riemannian manifold, which gives rise...

The Minimax Probability Machine (MPM) is an elegant machine learning algorithm for inductive learning. It learns a classifier that minimizes an upper bound on its own generalization error. In this paper , we extend its celebrated inductive formulation to an equally elegant transductive learning algorithm. In the transductive setting, the label assi...

The past years have witnessed many dedicated open-source projects that built and maintain implementations of Support Vector Machines (SVM), parallelized for GPU, multi-core CPUs and distributed systems. Up to this point, no comparable effort has been made to parallelize the Elastic Net, despite its popularity in many high impact applications, inclu...

A feature selection algorithm should ideally satisfy four conditions: reliably extract relevant features; be able to identify non-linear feature interactions; scale linearly with the number of features and dimensions; allow the incorporation of known sparsity structure. In this work we propose a novel feature selection algorithm, Gradient Boosted F...

In this paper, we propose a novel supervised learning method, Fast Flux Discriminant (FFD), for large-scale nonlinear classification. Compared with other existing methods, FFD has unmatched advantages, as it attains the efficiency and interpretability of linear models as well as the accuracy of nonlinear models. It is also sparse and naturally hand...

Bayesian optimization is a powerful framework for minimizing expensive objective functions while using very few function evaluations. It has been successfully applied to a variety of prob-lems, including hyperparameter tuning and ex-perimental design. However, this framework has not been extended to the inequality-constrained optimization setting,...

During the past decade, machine learning algorithms have become commonplace in large-scale real-world industrial applications. In these settings, the computation time to train and test machine learning algorithms is a key consideration. At training-time the algorithms must scale to very large data set sizes. At testing-time, the cost of feature ext...

Machine learning algorithms have successfully entered industry through many real-world applications (e.g. , search engines and product recommendations). In these applications, the test-time CPU cost must be budgeted and accounted for. In this paper, we examine two main components of the test-time CPU cost, classifier evaluation cost and feature ext...

In this paper, we evaluate the performance of various parallel optimization methods for Kernel Support Vector Machines on multicore CPUs and GPUs. In particular, we provide the first comparison of algorithms with explicit and implicit parallelization. Most existing parallel implementations for multi-core or GPU architectures are based on explicit p...

Nanoparticles are potentially powerful therapeutic tools that have the capacity to target drug payloads and imaging agents. However, some nanoparticles can activate complement, a branch of the innate immune system, and cause adverse side-effects. Recently, we employed an in vitro hemolysis assay to measure the serum complement activity of perfluoro...

The goal of machine learning is to develop predictors that generalize well to test data. Ideally, this is achieved by training on an almost infinitely large training data set that captures all variations in the data distribution. In practical learning settings, however, we do not have infinite data and our predictors may overfit. Overfitting may be...

We present Stochastic Neighbor Compression (SNC), an algorithm to compress a dataset for the purpose of fc-nearest neighbor (fcNN) classification. Given training data, SNC learns a much smaller synthetic data set, that minimizes the stochastic 1-nearest neighbor classification error on the training data. This approach has several appealing properti...

Denoising auto-encoders (DAEs) have been successfully used to learn new representations for a wide range of machine learning tasks. During training, DAEs make many passes over the train-ing dataset and reconstruct it from partial corruption generated from a pre-specified corrupting distribution. This process learns robust representation, though at...

Denoising auto-encoders (DAEs) have been successfully used to learn new representations for a wide range of machine learning tasks. During training, DAEs make many passes over the training dataset and reconstruct it from partial corruption generated from a pre-specified corrupting distribution. This process learns robust representation, though at t...

Glioblastoma Mulitforme is highly infiltrative, making precise delineation of tumor margin difficult. Multimodality or multi-parametric MR imaging sequences promise an advantage over anatomic sequences such as post contrast enhancement as methods for determining the spatial extent of tumor involvement. In considering multi-parametric imaging sequen...

Methods and system for rendering context aware multimedia content include identifying a plurality of multimedia content that is uploaded for rendering. The uploaded multimedia content is examined to determine metadata associated with each of the plurality of multimedia contents. Contextual information associated with the metadata is identified and...

In text mining, information retrieval, and machine learning, text documents are commonly represented through variants of sparse Bag of Words (sBoW) vectors (e.g. TF-IDF). Although simple and intuitive, sBoW style representations suffer from their inherent over-sparsity and fail to capture word-level synonymy and polysemy. Especially when labeled da...

An important problem in AI is to construct high-quality heuristics for optimal search. Recently, the Euclidean heuristic (EH) has been proposed, which embeds a state space graph into a Euclidean space and uses Euclidean distances as approximations for the graph distances. The embedding process leverages recent research re-sults from manifold learni...

Recently, a Euclidean heuristic (EH) has been proposed for A* search. EH exploits manifold learning methods to construct an embedding of the state space graph, and derives an admissible heuristic distance between two states from the Euclidean distance between their respective embedded points. EH has shown good performance and memory efficiency in c...

The goal of machine learning is to develop predictors that generalize well to test data. Ideally, this is achieved by training on very large (infinite) training data sets that capture all variations in the data distribution. In the case of finite training data, an effective solution is to extend the training set with artificially created examples -...

Evaluation cost during test-time is becoming increasingly important as many real-world applications need fast evaluation (e.g. web search engines, email spam filtering) or use expensive features (e.g. medical diagnosis). We introduce Anytime Feature Representations (AFR), a novel algorithm that explicitly addresses this trade-off in the data repres...

Automatic image annotation is a difficult and highly relevant machine learning task. Recent advances have significantly improved the state-of-the-art in retrieval accuracy with algorithms based on nearest neighbor classification in carefully learned metric spaces. But this comes at a price of increased computational complexity during training and t...

In this paper we introduce Maximum Variance Correction (MVC), which finds largescale feasible solutions to Maximum Variance Unfolding (MVU) by post-processing embeddings from any manifold learning algorithm. It increases the scale of MVU embeddings by several orders of magnitude and is naturally parallel. This unprecedented scalability opens up new...

In text mining, information retrieval, and machine learning, text documents are commonly represented through variants of sparse Bag of Words (sBoW) vectors (e.g. TF-IDF [1]). Although simple and intuitive, sBoW style representations suffer from their inherent over-sparsity and fail to capture word-level synonymy and polysemy. Especially when labele...

Recently, machine learning algorithms have successfully entered large-scale real-world industrial applications (e.g. search engines and email spam filters). Here, the CPU cost during test time must be budgeted and accounted for. In this paper, we address the challenge of balancing the test-time cost and the classifier accuracy in a principled fashi...

This paper considers the problem of learning an embedding of data based on similarity triplets of the form “A is more similar to B than to C”. This learning setting is of relevance to scenarios in which we wish to model human judgements on the similarity of objects. We argue that in order to obtain a truthful embedding of the underlying data, it is...

Recent work in metric learning has significantly improved the state-of-the-art in k-nearest neighbor classification. Support vector machines (SVM), particularly with RBF kernels, are amongst the most popular classification algorithms that uses distance metrics to compare examples. This paper provides an empirical analysis of the efficacy of three o...

As machine learning algorithms enter applications in industrial settings, there is increased interest in controlling their cpu-time during testing. The cpu-time consists of the running time of the algorithm and the extraction time of the features. The latter can vary drastically when the feature set is diverse. In this paper, we propose an algorith...

Stacked denoising autoencoders (SDAs) have been successfully used to learn new representations for domain adaptation. Recently, they have attained record accuracy on standard benchmark tasks of sentiment analysis across different text domains. SDAs learn robust data representations by reconstruction, recovering original features from data that are...

Machine learning algorithms are increasingly used in large-scale industrial settings. Here, the operational cost during test-time has to be taken into account when an algorithm is designed. This operational cost is affected by the average running time and the computation time required for feature extraction. When a diverse set of features is used,...

In this paper, we introduce two novel metric learning algorithms, χ2-LMNN and GB-LMNN, which are explicitly designed to be non-linear and easy-to-use. The two approaches achieve this goal in fundamentally different ways: χ2-LMNN inherits the computational benefits of a linear mapping from linear metric learning, but uses a non-linear χ2-distance to...

PURPOSE Glioblastoma multiforme (GBM) is the most common primary malignant neoplasm of the adult brain. The permeative nature of GBM makes tumor margin delineation challenging, at times limiting the efficacy of surgical resection. An improved method of assessing the geographic extent of GBM is therefore warranted. The authors hypothesize that a mac...

Web mail providers rely on users to "vote" to quickly and col-laboratively identify spam messages. Unfortunately, spammers have begun to use bots to control large collections of compromised Web mail accounts not just to send spam, but also to vote "not spam" on incoming spam emails in an attempt to thwart collaborative filtering. We call this pract...