Deep RL model learns lane following in 1 day.
problem Autonomous driving without explicit rules or supervision.
method Deep reinforcement learning, continuous model-free approach, single monocular input.
result Model learns lane following from random initialization in a few episodes.
Self-supervised method estimates depth from monocular endoscopy videos.
problem Depth estimation from monocular endoscopy data without manual labeling.
method Convolutional neural networks trained with sparse supervision from stereo methods.
result Submillimeter mean residual error in cross-patient CT scans comparison.
Improved self-supervised monocular depth estimation without complex architectures.
problem Challenges in acquiring per-pixel ground-truth depth data at scale.
method Proposed a set of improvements including a minimum reprojection loss, multi-scale sampling, and auto-masking loss.
result Surprisingly simple model leads to superior predictions compared to competing methods.
Self-supervised method estimates distances on fisheye cameras for autonomous driving.
problem Accurate Euclidean distance estimation on fisheye cameras for autonomous driving.
method Self-supervised scale-aware framework for monocular fisheye videos.
result State-of-the-art results on KITTI dataset, comparable to other methods.
Generates coherent 3D scenes from monocular videos without supervision.
problem Lack of 3D scene modeling in video generation models.
method Trains a model to generate 3D scenes with moving objects and a background from monocular videos.
result Trained model generates coherent 3D scenes with multiple moving objects and a background.
GANVO uses GANs to estimate camera motion and depth from unlabelled images.
problem Lack of labelled data for deep VO and depth estimation.
method Generative adversarial networks for unsupervised learning of 6-DoF pose and depth.
result Outperforms existing methods in pose estimation and depth recovery.
Regression Prior Networks improve ensemble performance on regression tasks.
problem Improving ensemble performance on regression tasks.
method Extending Prior Networks and Ensemble Distribution Distillation (EnD2) to regression tasks using the Normal-Wishart distribution. result Regression Prior Networks yield performance competitive with ensemble approaches on regression tasks.
Proposes a new metric to evaluate uncertainty in deep learning predictions.
problem Evaluating the reliability of deep learning predictions, especially uncertainty.
method Develops a novel metric for evaluating relative uncertainty in regression tasks with deep neural networks.
result Validates the new metric on a toy dataset and applies it to monocular depth estimation.
Learning based methods have shown very promising results for the task of depth estimation in single images. However, most existing approaches treat depth prediction as a supervised regression problem and as a result, require vast quantities of corresponding ground truth depth data for training. Just recording quality d…
Paper tackles depth estimation and optic disc-cup segmentation from color fundus images.
problem Depth estimation and optic disc-cup segmentation from color fundus images.
method Uses fully convolutional networks for monocular retinal depth estimation and optic disc-cup segmentation.
result Demonstrates improved accuracy in depth estimation and optic disc-cup segmentation.
Proposes a new method for robust uncertainty quantification in regression tasks.
problem Robust uncertainty estimation for deep neural networks in regression tasks.
method Generalized Auxiliary Uncertainty Estimator (AuxUE) scheme, considering both aleatoric and epistemic uncertainties.
result DIDO method provides robust uncertainty estimates in noisy inputs, scalable to image-level and pixel-wise tasks.
This paper improves 3D pose recovery from 2D images using non-convex regularization.
problem 3D object pose recovery from 2D images.
method Proposes non-convex regularization with leaky capped ℓ1-norm (LCNR) and multi-stage optimization.
result Theoretical analysis shows estimation error decreases with optimization stages.
We present a generalization of the Cauchy/Lorentzian, Geman-McClure, Welsch/Leclerc, generalized Charbonnier, Charbonnier/pseudo-Huber/L1-L2, and L2 loss functions. By introducing robustness as a continuous parameter, our loss function allows algorithms built around robust loss minimization to be generalized, which imp…
Enhances SSL methods with depth cues for better image understanding.
problem Lack of depth cues in 2D image pixel maps limits SSL performance.
method Integrates depth signals from a pretrained monocular RGB-to-depth model into contrastive learning frameworks.
result Improves SSL methods' robustness and generalization with depth signals.
NAVREN-RL uses deep reinforcement learning to teach drones indoor navigation.
problem Teaching drones to navigate indoor environments efficiently.
method End-to-end deep reinforcement learning with a custom reward function, expert data integration.
result Drones successfully navigate indoor arenas avoiding obstacles.
This paper explores the capabilities of convolutional neural networks to deal with a task that is easily manageable for humans: perceiving 3D pose of a human body from varying angles. However, in our approach, we are restricted to using a monocular vision system. For this purpose, we apply a convolutional neural networ…
New method improves reliability of depth estimation models.
problem Uncertainty quantification in large-scale vision models.
method Parameter-efficient Bayesian neural networks with PEFT methods.
result Combining PEFT methods with Bayesian inference enhances predictive performance.
We propose a method to generate multiple diverse and valid human pose hypotheses in 3D all consistent with the 2D detection of joints in a monocular RGB image. We use a novel generative model uniform (unbiased) in the space of anatomically plausible 3D poses. Our model is compositional (produces a pose by combining par…
Paper introduces a new regularization method for visual representations.
problem Learning sparse visual representations from over-complete data.
method Proposes leaky capped norm regularization (LCNR) and a majorization-minimization algorithm.
result LCNR outperforms ℓ1 regularization in monocular 3D shape recovery. New approach learns traffic behavior from unlabelled video.
problem Learning from demonstration in settings with unlabelled raw video data.
method Calibrates camera, detects objects, tracks them, and uses trajectories for LfD.
result Can learn traffic behavior from raw video without additional expert knowledge.
Y-GAN uses multi-camera data to estimate depth maps without expensive hardware.
problem Depth perception for autonomous systems requires accurate 3D spatial information.
method Proposes Y-GAN, a deep convolutional generative adversarial network.
result Y-GAN estimates depth maps from multi-camera stereo images without ground truth data.
End-to-end learning for robot grasping using image data.
problem Robotic grasping using only monocular images.
method Two-stream architecture: ventral stream for object detection and classification, dorsal stream for grasp planning.
result End-to-end trained model outperforms non-end-to-end systems.
Compact Gaussian model approximates deep ensemble predictions.
problem Efficiently approximating deep ensemble models for image prediction.
method Sparse-structured multivariate Gaussian with Cholesky parameterization trained to match pre-trained ensemble outputs.
result Compact representation captures uncertainty and structured correlations explicitly.
New method reconstructs 3D shapes from 2D images using Kendall's shape space.
problem Reconstruct 3D shapes from 2D images, especially for rare specimens.
method Kendall's shape space approach with prior information.
result More robust and plausible shapes compared to previous methods.
Hybrid RL combines simulated and real data for robust autonomous flight.
problem Challenges in training deep RL models for real-world robotic tasks.
method Combines real-world and simulated data to improve generalization.
result Quadrotor avoids collisions using only a monocular camera.
Aerial robot estimates human pose and path using dynamic classifier selection.
problem Estimating human pose and trajectory from aerial video.
method Dynamic classifier selection architecture; perspective correction; HOG and CNN features; 64 pose-viewpoint classes.
result Dynamic classifier selection improves efficiency and accuracy.
ETC improves Transformer models for long and structured inputs.
problem Scaling input length and encoding structured inputs in Transformers.
method Introduces global-local attention, relative position encodings, and CPC pre-training.
result Achieves state-of-the-art results on four natural language datasets.
Unified method for input, data, and model uncertainty in neural networks.
problem Uncertainty in neural network inputs and outputs.
method Propagating uncertainty through inputs using a unified formulation.
result More stable decision boundaries with input noise, and propagation of input uncertainty to model outputs.
New model evaluates how well models handle input faults.
problem Fault tolerance of models to input variations.
method Evaluates fault tolerance using information-based characteristic for arbitrary valid inputs.
result Proposes a new way to measure model robustness.
We reduce variance in RL with input-dependent baselines.
problem High variance in RL with standard baselines in input-driven environments.
method Derive and use a bias-free, input-dependent baseline; propose a meta-learning approach.
result Input-dependent baselines improve training stability and policy quality.
Improves sample efficiency in reinforcement learning with input representation.
problem Poor sample efficiency in reinforcement learning.
method Attention-based method to project inputs into an invariant representation space.
result Representation space is m! smaller for inputs of m objects, improving sample efficiency. Partial-input models fail to detect dataset artifacts, even when they perform poorly.
problem The effectiveness of partial-input models in detecting dataset artifacts is questionable.
method Design artificial datasets and identify trivial patterns in the SNLI dataset.
result Partial-input models can solve examples previously considered hard, indicating potential dataset artifacts.
New methods handle uncertainty in identifying input regions for a black-box function.
problem Handling uncertainty in identifying input regions for a black-box function.
method Introduce a basic framework and efficient methods for LSE under input uncertainty.
result Proposed methods can be applied to various LSE challenges under input uncertainty.
Detects unusual inputs to neural networks to prevent flawed predictions.
problem Erratic predictions from neural networks on unexpected inputs.
method Evaluates input unusualness by comparing its content to learned parameters.
result Simple, effective method for comparing input metrics across different scales.
Enhances construction input modeling with Bayesian deep neural networks.
problem Deriving reliable simulation input models from construction data.
method Bayesian deep neural networks integrated with multi-source construction data.
result Derives detailed input models for construction operations.
TVS-FNNs can approximate any continuous function on expanded input spaces.
problem Processing a broader range of inputs like sequences and matrices.
method Proving a universal approximation theorem for TVS-FNNs.
result TVS-FNNs can approximate any continuous function on expanded input spaces.
A new method builds sparse polynomial chaos expansions for models with dependent inputs.
problem Quantifying uncertainty in models with dependent inputs.
method Data-driven approach to construct orthonormal polynomials recursively based on input correlations.
result Reduces the number of observations and improves numerical stability and computational efficiency.
Improves AI system's understanding of human inputs by creating better examples.
problem AI systems misinterpret human inputs, leading to inefficiencies.
method Developed a conditional convolutional autoencoder (CCAE) to generate better examples.
result Generated examples lead to lower error rates and require less effort to create.
Validates neural networks inputs to protect against adversarial examples.
problem Ensuring neural networks robustness against adversarial attacks.
method Runtime local robustness verification based on normal distribution of robustness radii.
result Improves neural network accuracy and protects against adversarial examples.
Centroid Transformers reduce memory and computation by summarizing inputs into centroids.
problem Efficiently summarize inputs with reduced memory and computation.
method Generalizes self-attention to map N inputs to M centroids (M ≤ N), reducing complexity.
result Centroid Transformers reduce memory and computation while preserving key information.
MINs learn inverse mappings for high-dimensional optimization problems.
problem Data-driven optimization with high-dimensional inputs and valid subsets.
method Model Inversion Networks (MINs) learn an inverse mapping from scores to inputs.
result MINs can scale to high-dimensional input spaces and handle both offline and active data.
Methods for combining predictions from different models in a supervised learning setting must somehow estimate/predict the quality of a model's predictions at unknown future inputs. Many of these methods (often implicitly) make the assumption that the test inputs are identical to the training inputs, which is seldom re…
IA-BMA adapts model weights to inputs for better predictions.
problem Predicting with multiple models in heterogeneous settings.
method Input adaptive Bayesian Model Averaging (IA-BMA) with an input adaptive prior and amortized variational inference.
result IA-BMA consistently delivers more accurate and better-calibrated predictions.
Paper tackles machine performance testing under uncertain inputs.
problem Guarantee machine performance under input uncertainty.
method Formulates as IU-rLSE problem, proposes active learning method.
result Efficient algorithm for reliable level set estimation.
Framework verifies global correctness of neural networks for perception tasks.
problem Verifying robustness of neural networks is insufficient; global correctness needs to be ensured.
method Specified a state space and observation process to define the target input space. Tiled the spaces and compared ground truth and network output bounds to deliver error bounds.
result Framework can verify error bounds globally over the target input space and detect illegal inputs.
AIRBO optimizes robustly under uncertain inputs.
problem Challenges in Bayesian Optimization due to input uncertainty.
method AIRBO models arbitrary input uncertainty with MMD and accelerates posterior inference.
result AIRBO achieves state-of-the-art performance under various input uncertainties.
When simulating a complex stochastic system, the behavior of output response depends on input parameters estimated from finite real-world data, and the finiteness of data brings input uncertainty into the system. The quantification of the impact of input uncertainty on output response has been extensively studied. Most…
A new framework for differentially private ERM using input perturbation.
problem Ensuring privacy in empirical risk minimization with randomized data.
method Input perturbation where each data contributor independently randomizes their data.
result The model learned with input perturbation satisfies differential privacy and local differential privacy.