Bethge Lab

AnalysisBySynthesis

Adversarially Robust Neural Network on MNIST.

foolbox

A Python toolbox to create adversarial examples that fool neural networks in PyTorch, TensorFlow, and JAX

imagecorruptions

Python package to corrupt arbitrary images.

model-vs-human

Benchmark your model on out-of-distribution datasets with carefully collected human comparison data (NeurIPS 2021 Oral)

openimages2coco

Convert Open Images annotations into MS Coco format to make it a drop in replacement

robust-detection-benchmark

Code, data and benchmark from the paper "Benchmarking Robustness in Object Detection: Autonomous Driving when Winter is Coming" (NeurIPS 2019 ML4AD)

robustness

Robustness and adaptation of ImageNet scale models. Pre-Release, stay tuned for updates.

siamese-mask-rcnn

Siamese Mask R-CNN model for one-shot instance segmentation

slow_disentanglement

Towards Nonlinear Disentanglement in Natural Data with Temporal Sparse Coding

stylize-datasets

A script that applies the AdaIN style transfer method to arbitrary datasets

bethgelab/foolbox

A Python toolbox to create adversarial examples that fool neural networks in PyTorch, TensorFlow, and JAX

bethgelab/imagecorruptions

Python package to corrupt arbitrary images.

bethgelab/siamese-mask-rcnn

Siamese Mask R-CNN model for one-shot instance segmentation

bethgelab/model-vs-human

Benchmark your model on out-of-distribution datasets with carefully collected human comparison data (NeurIPS 2021 Oral)

bethgelab/robust-detection-benchmark

Code, data and benchmark from the paper "Benchmarking Robustness in Object Detection: Autonomous Driving when Winter is Coming" (NeurIPS 2019 ML4AD)

bethgelab/stylize-datasets

A script that applies the AdaIN style transfer method to arbitrary datasets

bethgelab/robustness

Robustness and adaptation of ImageNet scale models. Pre-Release, stay tuned for updates.

bethgelab/openimages2coco

Convert Open Images annotations into MS Coco format to make it a drop in replacement

bethgelab/slow_disentanglement

Towards Nonlinear Disentanglement in Natural Data with Temporal Sparse Coding

bethgelab/AnalysisBySynthesis

Adversarially Robust Neural Network on MNIST.

bethgelab/frequency_determines_performance

Code for the paper: "No Zero-Shot Without Exponential Data: Pretraining Concept Frequency Determines Multimodal Model Performance"

bethgelab/game-of-noise

Trained model weights, training and evaluation code from the paper "A simple way to make neural networks robust against diverse image corruptions"

bethgelab/InDomainGeneralizationBenchmark

bethgelab/slurm-monitoring-public

Monitor your high performance infrastructure configured over slurm using TIG stack

bethgelab/DataTypeIdentification

Code for the ICLR'24 paper: "Visual Data-Type Understanding does not emerge from Scaling Vision-Language Models"

bethgelab/magapi-wrapper

Wrapper around Microsoft Academic Knowledge API to retrieve MAG data

bethgelab/testing_visualizations

Code for the paper " Exemplary Natural Images Explain CNN Activations Better than Feature Visualizations"

bethgelab/docker-deeplearning

Development of new unified docker container (WIP)

bethgelab/notorious_difficulty_of_comparing_human_and_machine_perception

Code for the three case studies: Closed Contour Detection, Synthetic Visual Reasoning Test, Recognition Gap

bethgelab/lifelong-benchmarks

Benchmarks introduced in the paper: "Lifelong Benchmarks: Efficient Model Evaluation in an Era of Rapid Progress"

bethgelab/tools

bethgelab/mmdetection

Fork of the MMDetection Toolbox containing the Robustness Benchmark from the paper "Benchmarking Robustness in Object Detection: Autonomous Driving when Winter is Coming" (merged)

bethgelab/sort-and-search

Code for the paper: "Lifelong Benchmarks: Efficient Model Evaluation in an Era of Rapid Progress"

bethgelab/mnist_challenge

A challenge to explore adversarial robustness of neural networks on MNIST.

bethgelab/bwki-weekly-tasks

BWKI Task of the week

bethgelab/DeepLabCut

Markerless tracking of user-defined features with deep learning

bethgelab/defensive-distillation

bethgelab/foolbox-zoo-dummy

bethgelab/inference_results_v1.1

bethgelab/texture-vs-shape

Pre-trained models, data, code & materials from the paper "ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness" (ICLR 2019 Oral)