This work compares lattice-free and lattice-based training criteria for LVCSR.
problem Improving acoustic model performance in speech recognition.
method Direct comparison of lattice-free and lattice-based sequence discriminative training criteria using GPU.
result Lattice-free MMI performance is comparable to lattice-based criteria, while lattice-based sMBR remains superior.
Three LF training criteria improve neural network acoustic models without cross-entropy pre-training.
problem Improving purely sequence-trained neural network acoustic models.
method Comparison of three lattice-free discriminative training criteria (MMI, bMMI, sMBR) on LVCSR tasks.
result LF-bMMI models outperform plain LF-MMI models by 5% WER on Switchboard datasets.
Hybrid and end-to-end models compare in syllable recognition.
problem Comparing hybrid and end-to-end models for syllable recognition.
method Traditional hybrid system (kaldi) vs. end-to-end (TensorFlow) models.
result Hybrid models with explicit syllable knowledge outperform end-to-end models.
Taylorized training improves neural network training at finite width.
problem Understanding and improving neural network training at finite width.
method Training the k-th order Taylor expansion of the neural network at initialization.
result Taylorized training agrees with full neural network training better as k increases and can significantly close the performance gap.
Self-training outperforms pre-training on COCO object detection and segmentation datasets.
problem The effectiveness of pre-training in improving object detection and segmentation models is limited.
method Investigated self-training as an alternative method to utilize additional data.
result Self-training consistently improves model performance across various dataset sizes and data augmentation levels.
Un-trained neural networks outperform trained methods in MRI reconstruction.
problem Accelerated MRI reconstruction with minimal training data.
method Variation of Deep Decoder without training data.
result Un-trained approach significantly outperforms other methods in reconstruction accuracy.
Meta-learning performance depends on train-validation split type.
problem Understanding the importance of train-validation split in meta-learning.
method Theoretical and experimental study comparing train-val and train-train methods.
result Train-train method can achieve strictly better excess loss in realizable cases.
New method reveals how training data influence diffusion model outputs.
problem Difficulty in assessing training data impact on diffusion model outputs.
method Use of ensembles trained on carefully engineered splits of training data to identify influential training examples.
result Demonstrated the viability of ensembles as generative models and validity of assessing influence.
Paper shows adversarial training can be fooled by new type of noise.
problem Adversarial training can be fooled by new types of noise.
method Designing ADVIN, a new type of inducing noise.
result ADVIN can degrade adversarial training robustness by 99.9%.
New MIP methods improve training of integer-valued neural networks.
problem Training integer-valued neural networks with limited data and resources.
method Formulated new MIP models to optimize training efficiency and handle more data.
result Significantly outperforms previous state-of-the-art methods in accuracy, training time, and data usage.
Free adversarial training reduces the generalization gap compared to vanilla method.
problem Improving generalization in adversarial training.
method Analysis of algorithmic stability in free adversarial training.
result Free adversarial training shows a lower generalization gap.
ADASS selects adaptive subsets for SGD training acceleration.
problem Fixed sample size in SGD limits training efficiency.
method ADASS selects adaptive subsets based on Lipschitz constants.
result ADASS achieves comparable accuracy with full training set.
Solve-training trains neural nets to map physical solutions efficiently.
problem Representing complex physical solutions with neural networks.
method Variational training using loss functions from physical models.
result Effective neural network representation of solution maps without expensive labels.
MixTrain improves verifiable robustness of neural networks without sacrificing efficiency.
problem Efficiently making neural networks robust against adversarial attacks.
method Stochastic robust approximation and dynamic mixed training techniques.
result Achieves up to 95.2% verified robust accuracy with significantly reduced training time.
learn2mix trains neural nets faster by adjusting class proportions dynamically.
problem Training neural nets efficiently with limited resources and imbalanced classes.
method Adaptive class proportion adjustment during training.
result Neural nets trained with learn2mix converge faster than static methods.
Loss-guided training accelerates node embedding methods on graphs.
problem Training efficiency in graph learning methods with implicit positive examples.
method Dynamic adjustment of training distribution based on loss values.
result Significant acceleration in training and computation over static methods.
Paper explores fast adversarial training to improve robustness with less computation.
problem Efficiently defending against adversarial examples.
method Integrates simple self-attacks for faster training, focusing on overfitting recovery.
result Shows superior robust accuracy with reduced training time compared to strong adversarial training.
Free adversarial training improves robustness without generating adversarial examples.
problem Training robust models against adversarial attacks is costly and impractical for large-scale datasets.
method Recycles gradient information from parameter updates to generate adversarial examples.
result Free adversarial training achieves comparable robustness to PGD training at negligible cost.
Detects backdoors in trained models without poisoned training data.
problem Detecting backdoors in DNNs trained without access to the poisoned training set.
method Proposes a novel detector using the maximum achievable misclassification fraction (MAMF) statistic.
result Detects backdoors and infers source and target classes.
Pre-training improves model robustness and uncertainty.
problem Improving model robustness and uncertainty in machine learning.
method Adversarial pre-training and task-specific methods.
result Approximately a 10% absolute improvement in adversarial robustness.
Crowdsourced training of large neural networks with decentralized Mixture-of-Experts.
problem Expensive training of large neural networks limits research contributions.
method Learning@home: decentralized Mixture-of-Experts for large, poorly connected participants.
result Performance and reliability of Learning@home surpass conventional distributed training.
Pre-training is crucial for learning deep neural networks. Most of existing pre-training methods train simple models (e.g., restricted Boltzmann machines) and then stack them layer by layer to form the deep structure. This layer-wise pre-training has found strong theoretical foundation and broad empirical support. Howe…
New method reconstructs significant parts of training data from neural networks.
problem Understanding and reconstructing training data from neural networks.
method Proposes a novel reconstruction scheme based on recent theoretical results about neural network training.
result Shows that a significant fraction of training data can be reconstructed from neural network parameters.
Reduces deep learning training data for faster testing.
problem Resource-intensive deep learning training with full data sets.
method Evaluated different training set reduction methods.
result Training set reduction is useful in resource-constrained environments.
The paper studies the asymptotic behavior of adversarial training under ℓ∞-perturbation.
problem Theoretical guarantees for sparsity-recovery in adversarial training.
method Investigation of the asymptotic distribution of the adversarial training estimator in generalized linear models.
result The asymptotic distribution of the adversarial training estimator under ℓ∞-perturbation could have a positive probability mass at 0 when the true parameter is 0. This paper surveys distributed training techniques for deep learning models.
problem Substantial compute needed for training deep learning models.
method Exploration of various algorithms and techniques for distributed training.
result Recent advancements have reduced training time from weeks to minutes.
This paper introduces 'General Cyclical Training' for neural networks.
problem Improving training efficiency and performance of neural networks.
method Cyclical training phases with varying hyperparameters, batch sizes, loss functions, and data augmentation.
result Cyclical weight decay, softmax temperature, and gradient clipping enhance model accuracy.
Speeds up training and inference by pruning entire channels before training.
problem Training and inference speed in deep neural networks.
method Structured pruning applied before training, focusing on removing entire channels and hidden units.
result 2x speedup in training and 3x speedup in inference.
Improved GAN training with defined weights speeds up convergence and accuracy.
problem Challenges in training Generative Adversarial Networks (GANs).
method Proposed a training variation with defined weights to assist GAN training.
result The new algorithm improves GAN performance by 5% to 50% in synthetic and image datasets.
The paper predicts loss scaling across different datasets and compute scales.
problem Predicting loss scaling across different datasets and compute scales.
method Derive shifted power law relationships between train and test losses.
result Shifted power law relationships hold for various datasets and tasks, improving prediction accuracy.
Theoretical study shows adversarial training improves robustness in deep learning models.
problem Ensuring robustness in pre-trained deep learning models.
method Theoretical analysis of adversarial training and feature purification in two-layer neural networks.
result Adversarial training leads to feature purification, making models more robust to attacks.
Training dynamics reveal neural network complexity.
problem Understanding the complexity of neural networks during training.
method Relating the Lipschitz constant and training behavior of neural networks.
result Steady training and bounded complexity in far regions of input space.
LIIT uses large deviations to train neural networks faster.
problem Training neural networks requires large amounts of data, leading to inefficiencies.
method LIIT uses large deviations principle to generate and update training samples.
result LIIT achieves similar classification performance with smaller, more representative samples.
Empirical study shows GANs overfit and drop modes when training is deterministic.
problem Understanding overfitting and mode drop in GAN training.
method Empirical analysis of GAN training with and without stochasticity.
result GANs overfit and drop modes when training is deterministic.
Joint training improves model accuracy by selectively using privileged information.
problem Two-stage training can lead to model failure with noisy privileged information.
method Joint training of two models to use privileged information selectively.
result Joint training outperforms two-stage baselines on synthetic and real-world tasks.
ShadowSync separates background synchronization for scalable distributed training.
problem Reducing synchronization overhead in distributed training for high scalability.
method Separates synchronization from training and runs it in the background.
result Achieves both high throughput and excellent model quality at scale.
New strategy improves GNN pre-training for graph datasets.
problem Effective pre-training for graph neural networks on graph datasets.
method Expressive pre-training of GNNs at both node and graph levels.
result Significant improvement in generalization across downstream tasks.
MACER trains robust models without adversarial training, faster and more effective.
problem Learning robust models without relying on attack-dependent adversarial training.
method MACER trains provably robust smoothed classifiers by maximizing certified radius.
result MACER achieves larger average certified radius and faster training time compared to state-of-the-art methods.
The study finds a theoretical bound for pre-training iterations needed for pruning to yield good subnetwork performance.
problem Discovering efficient subnetworks within pre-trained dense networks.
method Mathematical analysis of a two-layer, fully-connected network, validating with a multi-layer perceptron trained on MNIST.
result A logarithmically dependent threshold on dataset size for successful pruning.
New method improves adversarial training efficiency and robustness.
problem High computational costs and lack of stability in adversarial training.
method Backward smoothing for randomized smoothing of random initialization.
result Our method achieves similar model robustness as state-of-the-art methods but with significantly less training time.
Recent work on discriminative segmental models has shown that they can achieve competitive speech recognition performance, using features based on deep neural frame classifiers. However, segmental models can be more challenging to train than standard frame-based approaches. While some segmental models have been success…
Improved GAN training speed and quality with FastGAN.
problem Slower convergence and less expressive models in GAN training.
method Adversarial training with FastGAN algorithm.
result Better generation quality with less training time.
DES training speeds up large-scale recommender systems convergence.
problem Training large-scale recommender systems with dynamic sparse features.
method Distributed Equivalent Substitution (DES) framework for fully synchronous training.
result DES achieves higher AUC and up to 68.7% communication savings.
Predict training time without training using a linearized model.
problem Predicting the training time of a deep network without actual training.
method Approximate training dynamics with a linearized model and solve a low-dimensional SDE in function space.
result Predict training time within 20% error margin and 30-45x reduction in cost.
We propose a max-pooling based loss function for training Long Short-Term Memory (LSTM) networks for small-footprint keyword spotting (KWS), with low CPU, memory, and latency requirements. The max-pooling loss training can be further guided by initializing with a cross-entropy loss trained network. A posterior smoothin…
GP-TS optimizes TLM pre-training hyperparameters efficiently.
problem Resource inefficiency in TLM pre-training.
method Bayesian optimization with Thompson sampling and Gaussian process.
result GP-TS achieves lower MLM loss in fewer epochs.
Study shows E2E training fails for over-parameterized models.
problem When does E2E training fail for complex Deep Network architectures?
method Blend gradient between independent training and E2E training.
result Optimum can lie between ensemble and E2E, challenging traditional approaches.
New unsupervised neural network for beamforming masks trained on real noisy speech.
problem Training neural beamforming masks without labeled data.
method Unsupervised training using a likelihood criterion from spatial mixture model.
result Performance comparable to supervised and teacher-trained systems.