arXiv subjects

Oncel Tuzel

Publications and source records attributed to Oncel Tuzel.

At least 73 records · Page 4Linked to original sources

arXiv2021

Token Pooling in Vision Transformers

Despite the recent success in many applications, the high computational requirements of vision transformers limit their use in resource-constrained settings. While many existing methods improve the quadratic complexity of attention, in most vision transformers, self-attention is not the major computation bottleneck, e.g., more than 80% of the computation is spent on fully-connected layers. To improve the computational complexity of all layers, we propose a novel token downsampling method, called Token Pooling, efficiently exploiting redundancies in the images and intermediate token representations. We show that, under mild assumptions, softmax-attention acts as a high-dimensional low-pass (smoothing) filter. Thus, its output contains redundancy that can be pruned to achieve a better trade-off between the computational cost and accuracy. Our new technique accurately approximates a set of tokens by minimizing the reconstruction error caused by downsampling. We solve this optimization problem via cost-efficient clustering. We rigorously analyze and compare to prior downsampling methods. Our experiments show that Token Pooling significantly improves the cost-accuracy trade-off over the state-of-the-art downsampling. Token Pooling is a simple and effective operator that can benefit many architectures. Applied to DeiT, it achieves the same ImageNet top-1 accuracy using 42% fewer computations.

cs.CV↗

arXiv2021

Instance-Level Task Parameters: A Robust Multi-task Weighting Framework

Recent works have shown that deep neural networks benefit from multi-task learning by learning a shared representation across several related tasks. However, performance of such systems depend on relative weighting between various losses involved during training. Prior works on loss weighting schemes assume that instances are equally easy or hard for all tasks. In order to break this assumption, we let the training process dictate the optimal weighting of tasks for every instance in the dataset. More specifically, we equip every instance in the dataset with a set of learnable parameters (instance-level task parameters) where the cardinality is equal to the number of tasks learned by the model. These parameters model the weighting of each task for an instance. They are updated by gradient descent and do not require hand-crafted rules. We conduct extensive experiments on SURREAL and CityScapes datasets, for human shape and pose estimation, depth estimation and semantic segmentation tasks. In these tasks, our approach outperforms recent dynamic loss weighting approaches, e.g. reducing surface estimation errors by 8.97% on SURREAL. When applied to datasets where one or more tasks can have noisy annotations, the proposed method learns to prioritize learning from clean labels for a given task, e.g. reducing surface estimation errors by up to 60%. We also show that we can reliably detect corrupt labels for a given task as a by-product from learned instance-level task parameters.

cs.CV↗

arXiv2021

Optimize what matters: Training DNN-HMM Keyword Spotting Model Using End Metric

Deep Neural Network--Hidden Markov Model (DNN-HMM) based methods have been successfully used for many always-on keyword spotting algorithms that detect a wake word to trigger a device. The DNN predicts the state probabilities of a given speech frame, while HMM decoder combines the DNN predictions of multiple speech frames to compute the keyword detection score. The DNN, in prior methods, is trained independent of the HMM parameters to minimize the cross-entropy loss between the predicted and the ground-truth state probabilities. The mis-match between the DNN training loss (cross-entropy) and the end metric (detection score) is the main source of sub-optimal performance for the keyword spotting task. We address this loss-metric mismatch with a novel end-to-end training strategy that learns the DNN parameters by optimizing for the detection score. To this end, we make the HMM decoder (dynamic programming) differentiable and back-propagate through it to maximize the score for the keyword and minimize the scores for non-keyword speech segments. Our method does not require any change in the model architecture or the inference framework; therefore, there is no overhead in run-time memory or compute requirements. Moreover, we show significant reduction in false rejection rate (FRR) at the same false trigger experience (> 70% over independent DNN training).

cs.SD↗

arXiv2021

SapAugment: Learning A Sample Adaptive Policy for Data Augmentation

Data augmentation methods usually apply the same augmentation (or a mix of them) to all the training samples. For example, to perturb data with noise, the noise is sampled from a Normal distribution with a fixed standard deviation, for all samples. We hypothesize that a hard sample with high training loss already provides strong training signal to update the model parameters and should be perturbed with mild or no augmentation. Perturbing a hard sample with a strong augmentation may also make it too hard to learn from. Furthermore, a sample with low training loss should be perturbed by a stronger augmentation to provide more robustness to a variety of conditions. To formalize these intuitions, we propose a novel method to learn a Sample-Adaptive Policy for Augmentation -- SapAugment. Our policy adapts the augmentation parameters based on the training loss of the data samples. In the example of Gaussian noise, a hard sample will be perturbed with a low variance noise and an easy sample with a high variance noise. Furthermore, the proposed method combines multiple augmentation methods into a methodical policy learning framework and obviates hand-crafting augmentation parameters by trial-and-error. We apply our method on an automatic speech recognition (ASR) task, and combine existing and novel augmentations using the proposed framework. We show substantial improvement, up to 21% relative reduction in word error rate on LibriSpeech dataset, over the state-of-the-art speech augmentation method.

cs.LG↗

arXiv2020

Extracurricular Learning: Knowledge Transfer Beyond Empirical Distribution

Knowledge distillation has been used to transfer knowledge learned by a sophisticated model (teacher) to a simpler model (student). This technique is widely used to compress model complexity. However, in most applications the compressed student model suffers from an accuracy gap with its teacher. We propose extracurricular learning, a novel knowledge distillation method, that bridges this gap by (1) modeling student and teacher output distributions; (2) sampling examples from an approximation to the underlying data distribution; and (3) matching student and teacher output distributions over this extended set including uncertain samples. We conduct rigorous evaluations on regression and classification tasks and show that compared to the standard knowledge distillation, extracurricular learning reduces the gap by 46% to 68%. This leads to major accuracy improvements compared to the empirical risk minimization-based training for various recent neural network architectures: 16% regression error reduction on the MPIIGaze dataset, +3.4% to +9.1% improvement in top-1 classification accuracy on the CIFAR100 dataset, and +2.9% top-1 improvement on the ImageNet dataset.

cs.LG↗

arXiv2020

Subject-Aware Contrastive Learning for Biosignals

Datasets for biosignals, such as electroencephalogram (EEG) and electrocardiogram (ECG), often have noisy labels and have limited number of subjects (<100). To handle these challenges, we propose a self-supervised approach based on contrastive learning to model biosignals with a reduced reliance on labeled data and with fewer subjects. In this regime of limited labels and subjects, intersubject variability negatively impacts model performance. Thus, we introduce subject-aware learning through (1) a subject-specific contrastive loss, and (2) an adversarial training to promote subject-invariance during the self-supervised learning. We also develop a number of time-series data augmentation techniques to be used with the contrastive loss for biosignals. Our method is evaluated on publicly available datasets of two different biosignals with different tasks: EEG decoding and ECG anomaly detection. The embeddings learned using self-supervision yield competitive classification results compared to entirely supervised methods. We show that subject-invariance improves representation quality for these tasks, and observe that subject-specific loss increases performance when fine-tuning with supervised labels.

cs.LG↗

arXiv2020

Least squares binary quantization of neural networks

Quantizing weights and activations of deep neural networks results in significant improvement in inference efficiency at the cost of lower accuracy. A source of the accuracy gap between full precision and quantized models is the quantization error. In this work, we focus on the binary quantization, in which values are mapped to -1 and 1. We provide a unified framework to analyze different scaling strategies. Inspired by the pareto-optimality of 2-bits versus 1-bit quantization, we introduce a novel 2-bits quantization with provably least squares error. Our quantization algorithms can be implemented efficiently on the hardware using bitwise operations. We present proofs to show that our proposed methods are optimal, and also provide empirical error analysis. We conduct experiments on the ImageNet dataset and show a reduced accuracy gap when using the proposed least squares quantization algorithms.

cs.LG↗

arXiv2020

Unsupervised Style and Content Separation by Minimizing Mutual Information for Speech Synthesis

We present a method to generate speech from input text and a style vector that is extracted from a reference speech signal in an unsupervised manner, i.e., no style annotation, such as speaker information, is required. Existing unsupervised methods, during training, generate speech by computing style from the corresponding ground truth sample and use a decoder to combine the style vector with the input text. Training the model in such a way leaks content information into the style vector. The decoder can use the leaked content and ignore some of the input text to minimize the reconstruction loss. At inference time, when the reference speech does not match the content input, the output may not contain all of the content of the input text. We refer to this problem as "content leakage", which we address by explicitly estimating and minimizing the mutual information between the style and the content through an adversarial training formulation. We call our method MIST - Mutual Information based Style Content Separation. The main goal of the method is to preserve the input content in the synthesized speech signal, which we measure by the word error rate (WER) and show substantial improvements over state-of-the-art unsupervised speech synthesis methods.

eess.AS↗

arXiv2019

Nonlinear Conjugate Gradients For Scaling Synchronous Distributed DNN Training

Nonlinear conjugate gradient (NLCG) based optimizers have shown superior loss convergence properties compared to gradient descent based optimizers for traditional optimization problems. However, in Deep Neural Network (DNN) training, the dominant optimization algorithm of choice is still Stochastic Gradient Descent (SGD) and its variants. In this work, we propose and evaluate the stochastic preconditioned nonlinear conjugate gradient algorithm for large scale DNN training tasks. We show that a nonlinear conjugate gradient algorithm improves the convergence speed of DNN training, especially in the large mini-batch scenario, which is essential for scaling synchronous distributed DNN training to large number of workers. We show how to efficiently use second order information in the NLCG pre-conditioner for improving DNN training convergence. For the ImageNet classification task, at extremely large mini-batch sizes of greater than 65k, NLCG optimizer is able to improve top-1 accuracy by more than 10 percentage points for standard training of the Resnet-50 model for 90 epochs. For the CIFAR-100 classification task, at extremely large mini-batch sizes of greater than 16k, NLCG optimizer is able to improve top-1 accuracy by more than 15 percentage points for standard training of the Resnet-32 model for 200 epochs.

cs.LG↗

arXiv2019

Divide, Denoise, and Defend against Adversarial Attacks

Deep neural networks, although shown to be a successful class of machine learning algorithms, are known to be extremely unstable to adversarial perturbations. Improving the robustness of neural networks against these attacks is important, especially for security-critical applications. To defend against such attacks, we propose dividing the input image into multiple patches, denoising each patch independently, and reconstructing the image, without losing significant image content. We call our method D3. This proposed defense mechanism is non-differentiable which makes it non-trivial for an adversary to apply gradient-based attacks. Moreover, we do not fine-tune the network with adversarial examples, making it more robust against unknown attacks. We present an analysis of the tradeoff between accuracy and robustness against adversarial attacks. We evaluate our method under black-box, grey-box, and white-box settings. On the ImageNet dataset, our method outperforms the state-of-the-art by 19.7% under grey-box setting, and performs comparably under black-box setting. For the white-box setting, the proposed method achieves 34.4% accuracy compared to the 0% reported in the recent works.

cs.CV↗

arXiv2019

MVX-Net: Multimodal VoxelNet for 3D Object Detection

Many recent works on 3D object detection have focused on designing neural network architectures that can consume point cloud data. While these approaches demonstrate encouraging performance, they are typically based on a single modality and are unable to leverage information from other modalities, such as a camera. Although a few approaches fuse data from different modalities, these methods either use a complicated pipeline to process the modalities sequentially, or perform late-fusion and are unable to learn interaction between different modalities at early stages. In this work, we present PointFusion and VoxelFusion: two simple yet effective early-fusion approaches to combine the RGB and point cloud modalities, by leveraging the recently introduced VoxelNet architecture. Evaluation on the KITTI dataset demonstrates significant improvements in performance over approaches which only use point cloud data. Furthermore, the proposed method provides results competitive with the state-of-the-art multimodal algorithms, achieving top-2 ranking in five of the six bird's eye view and 3D detection categories on the KITTI benchmark, by using a simple single stage network.

cs.CV↗

arXiv2017

VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection

Accurate detection of objects in 3D point clouds is a central problem in many applications, such as autonomous navigation, housekeeping robots, and augmented/virtual reality. To interface a highly sparse LiDAR point cloud with a region proposal network (RPN), most existing efforts have focused on hand-crafted feature representations, for example, a bird's eye view projection. In this work, we remove the need of manual feature engineering for 3D point clouds and propose VoxelNet, a generic 3D detection network that unifies feature extraction and bounding box prediction into a single stage, end-to-end trainable deep network. Specifically, VoxelNet divides a point cloud into equally spaced 3D voxels and transforms a group of points within each voxel into a unified feature representation through the newly introduced voxel feature encoding (VFE) layer. In this way, the point cloud is encoded as a descriptive volumetric representation, which is then connected to a RPN to generate detections. Experiments on the KITTI car detection benchmark show that VoxelNet outperforms the state-of-the-art LiDAR based 3D detection methods by a large margin. Furthermore, our network learns an effective discriminative representation of objects with various geometries, leading to encouraging results in 3D detection of pedestrians and cyclists, based on only LiDAR.

cs.CV↗

arXiv2017

Learning from Simulated and Unsupervised Images through Adversarial Training

With recent progress in graphics, it has become more tractable to train models on synthetic images, potentially avoiding the need for expensive annotations. However, learning from synthetic images may not achieve the desired performance due to a gap between synthetic and real image distributions. To reduce this gap, we propose Simulated+Unsupervised (S+U) learning, where the task is to learn a model to improve the realism of a simulator's output using unlabeled real data, while preserving the annotation information from the simulator. We develop a method for S+U learning that uses an adversarial network similar to Generative Adversarial Networks (GANs), but with synthetic images as inputs instead of random vectors. We make several key modifications to the standard GAN algorithm to preserve annotations, avoid artifacts, and stabilize training: (i) a 'self-regularization' term, (ii) a local adversarial loss, and (iii) updating the discriminator using a history of refined images. We show that this enables generation of highly realistic images, which we demonstrate both qualitatively and with a user study. We quantitatively evaluate the generated images by training models for gaze estimation and hand pose estimation. We show a significant improvement over using synthetic images, and achieve state-of-the-art results on the MPIIGaze dataset without any labeled real data.

cs.CV↗

arXiv2017

Attentional Network for Visual Object Detection

We propose augmenting deep neural networks with an attention mechanism for the visual object detection task. As perceiving a scene, humans have the capability of multiple fixation points, each attended to scene content at different locations and scales. However, such a mechanism is missing in the current state-of-the-art visual object detection methods. Inspired by the human vision system, we propose a novel deep network architecture that imitates this attention mechanism. As detecting objects in an image, the network adaptively places a sequence of glimpses of different shapes at different locations in the image. Evidences of the presence of an object and its location are extracted from these glimpses, which are then fused for estimating the object class and bounding box coordinates. Due to lacks of ground truth annotations of the visual attention mechanism, we train our network using a reinforcement learning algorithm with policy gradients. Experiment results on standard object detection benchmarks show that the proposed network consistently outperforms the baseline networks that does not model the attention mechanism.

cs.CV↗

arXiv2016

Robust Face Alignment Using a Mixture of Invariant Experts

Face alignment, which is the task of finding the locations of a set of facial landmark points in an image of a face, is useful in widespread application areas. Face alignment is particularly challenging when there are large variations in pose (in-plane and out-of-plane rotations) and facial expression. To address this issue, we propose a cascade in which each stage consists of a mixture of regression experts. Each expert learns a customized regression model that is specialized to a different subset of the joint space of pose and expressions. The system is invariant to a predefined class of transformations (e.g., affine), because the input is transformed to match each expert's prototype shape before the regression is applied. We also present a method to include deformation constraints within the discriminative alignment framework, which makes our algorithm more robust. Our algorithm significantly outperforms previous methods on publicly available face alignment datasets.

cs.CV↗

arXiv2016

Coupled Generative Adversarial Networks

We propose coupled generative adversarial network (CoGAN) for learning a joint distribution of multi-domain images. In contrast to the existing approaches, which require tuples of corresponding images in different domains in the training set, CoGAN can learn a joint distribution without any tuple of corresponding images. It can learn a joint distribution with just samples drawn from the marginal distributions. This is achieved by enforcing a weight-sharing constraint that limits the network capacity and favors a joint distribution solution over a product of marginal distributions one. We apply CoGAN to several joint distribution learning tasks, including learning a joint distribution of color and depth images, and learning a joint distribution of face images with different attributes. For each task it successfully learns the joint distribution without any tuple of corresponding images. We also demonstrate its applications to domain adaptation and image transformation.

cs.CV↗

arXiv2016

Global-Local Face Upsampling Network

Face hallucination, which is the task of generating a high-resolution face image from a low-resolution input image, is a well-studied problem that is useful in widespread application areas. Face hallucination is particularly challenging when the input face resolution is very low (e.g., 10 x 12 pixels) and/or the image is captured in an uncontrolled setting with large pose and illumination variations. In this paper, we revisit the algorithm introduced in [1] and present a deep interpretation of this framework that achieves state-of-the-art under such challenging scenarios. In our deep network architecture the global and local constraints that define a face can be efficiently modeled and learned end-to-end using training data. Conceptually our network design can be partitioned into two sub-networks: the first one implements the holistic face reconstruction according to global constraints, and the second one enhances face-specific details and enforces local patch statistics. We optimize the deep network using a new loss function for super-resolution that combines reconstruction error with a learned face quality measure in adversarial setting, producing improved visual results. We conduct extensive experiments in both controlled and uncontrolled setups and show that our algorithm improves the state of the art both numerically and visually.

cs.CV↗

arXiv2015

Deep Gaussian Conditional Random Field Network: A Model-based Deep Network for Discriminative Denoising

We propose a novel deep network architecture for image\\ denoising based on a Gaussian Conditional Random Field (GCRF) model. In contrast to the existing discriminative denoising methods that train a separate model for each noise level, the proposed deep network explicitly models the input noise variance and hence is capable of handling a range of noise levels. Our deep network, which we refer to as deep GCRF network, consists of two sub-networks: (i) a parameter generation network that generates the pairwise potential parameters based on the noisy input image, and (ii) an inference network whose layers perform the computations involved in an iterative GCRF inference procedure.\ We train the entire deep GCRF network (both parameter generation and inference networks) discriminatively in an end-to-end fashion by maximizing the peak signal-to-noise ratio measure. Experiments on Berkeley segmentation and PASCALVOC datasets show that the proposed deep GCRF network outperforms state-of-the-art image denoising approaches for several noise levels.

cs.CV↗