Philipp Berens
@philipp.hertie.ai
Professor of Data Science @ University of Tübingen, Director of Hertie AI (www.hertie.ai) and Speaker of ML4Science (www.machinelearningforscience.de)
Stunning 18th place in finish! @ml4science.bsky.social
June 28, 2025 at 2:47 PM
Stunning 18th place in finish! @ml4science.bsky.social
The colleagues from the new excellence cluster TERRA @unituebingen.bsky.social arriving after 100k on the 9th place!
June 28, 2025 at 2:37 PM
The colleagues from the new excellence cluster TERRA @unituebingen.bsky.social arriving after 100k on the 9th place!
@ml4science.bsky.social is currently 17th in the 100k relay race of @unituebingen.bsky.social! 14 laps to go! #ai4science
June 28, 2025 at 1:47 PM
@ml4science.bsky.social is currently 17th in the 100k relay race of @unituebingen.bsky.social! 14 laps to go! #ai4science
There should be a categorical imperative for bikes: Treat your bike like you would want to be treated by it
May 26, 2025 at 7:38 AM
There should be a categorical imperative for bikes: Treat your bike like you would want to be treated by it
This is where @hertie-ai.bsky.social will be located in three years or so!
May 22, 2025 at 2:24 PM
This is where @hertie-ai.bsky.social will be located in three years or so!
Importantly, we do not simply compute quality metrics, but perform a highly sensitive three-way odd-one-out study to assess the realism - even for ophthalmologists, it is hard to figure out, which images have been generated by our algorithm!
May 22, 2025 at 7:28 AM
Importantly, we do not simply compute quality metrics, but perform a highly sensitive three-way odd-one-out study to assess the realism - even for ophthalmologists, it is hard to figure out, which images have been generated by our algorithm!
This allows to generate high quality counterfactual images for individual patients, e.g. answering the question how an image had looked if the patient had been healthy or if she had been sick. The induced lesions look highly realistic, and the vessel structure is kept intact.
May 22, 2025 at 7:28 AM
This allows to generate high quality counterfactual images for individual patients, e.g. answering the question how an image had looked if the patient had been healthy or if she had been sick. The induced lesions look highly realistic, and the vessel structure is kept intact.
In our model, we use the gradients of a robust disease classifier and regularization towards the original image to guide the diffusion process
May 22, 2025 at 7:28 AM
In our model, we use the gradients of a robust disease classifier and regularization towards the original image to guide the diffusion process
And another one: We show how to use a diffusion model for #fundus images for #counterfactual reasoning in #ophthalmology! If you ever wondered how the fundus images of your patient had looked, if he had been suffering from diabetic retinopathy, check this out ⬇️
journals.plos.org/digitalhealt...
journals.plos.org/digitalhealt...
May 22, 2025 at 7:28 AM
And another one: We show how to use a diffusion model for #fundus images for #counterfactual reasoning in #ophthalmology! If you ever wondered how the fundus images of your patient had looked, if he had been suffering from diabetic retinopathy, check this out ⬇️
journals.plos.org/digitalhealt...
journals.plos.org/digitalhealt...
This not only works much better than established post-hoc techniques, but also works well as a support tool for clinicians: Decisions for difficult cases improve and all decisions become about 20% faster!
May 13, 2025 at 7:04 AM
This not only works much better than established post-hoc techniques, but also works well as a support tool for clinicians: Decisions for difficult cases improve and all decisions become about 20% faster!
The explicit class evidence maps allows to penalize activations, e.g. via a sparseness penalty. This is extremely effective for diseases like #diabetic #retinopathy where lesions are small in initial disease stages: extracted high evidence regions almost always contain lesions.
May 13, 2025 at 7:04 AM
The explicit class evidence maps allows to penalize activations, e.g. via a sparseness penalty. This is extremely effective for diseases like #diabetic #retinopathy where lesions are small in initial disease stages: extracted high evidence regions almost always contain lesions.
BagNets are modified ResNets with local receptive fields and an explicit class evidence maps at the end. We discussed there use for interpretable #medical #AI here: proceedings.mlr.press/v227/donteu2...
May 13, 2025 at 6:59 AM
BagNets are modified ResNets with local receptive fields and an explicit class evidence maps at the end. We discussed there use for interpretable #medical #AI here: proceedings.mlr.press/v227/donteu2...
We are excited that our work on inherently interpretable #deeplearning models for #AI for #medicine has been published in @plos.org #digitalhealth! You want to know how to combine the power of deep learning with accessible interpretation? This is for you! ⬇️ journals.plos.org/digitalhealt...
May 13, 2025 at 6:53 AM
We are excited that our work on inherently interpretable #deeplearning models for #AI for #medicine has been published in @plos.org #digitalhealth! You want to know how to combine the power of deep learning with accessible interpretation? This is for you! ⬇️ journals.plos.org/digitalhealt...
A new use case for ChatGPT: Make grants and other scientific texts adhere to US-style censoring.
May 3, 2025 at 4:44 PM
A new use case for ChatGPT: Make grants and other scientific texts adhere to US-style censoring.
Ah, ChatGPT comes to your help:
May 3, 2025 at 4:37 PM
Ah, ChatGPT comes to your help:
Are you looking for a #PhD or #Postdoc in #computational #neuroscience and #machinelearning for developing new models for testing theories of efficient coding? Join our team @hertie-ai.bsky.social at @unituebingen.bsky.social @ml4science.bsky.social!
jobs.medizin.uni-tuebingen.de/Job/5996/Pos...
jobs.medizin.uni-tuebingen.de/Job/5996/Pos...
March 14, 2025 at 2:45 PM
Are you looking for a #PhD or #Postdoc in #computational #neuroscience and #machinelearning for developing new models for testing theories of efficient coding? Join our team @hertie-ai.bsky.social at @unituebingen.bsky.social @ml4science.bsky.social!
jobs.medizin.uni-tuebingen.de/Job/5996/Pos...
jobs.medizin.uni-tuebingen.de/Job/5996/Pos...
Are you looking for a #PhD or #Postdoc in #bioinformatics and #machinelearning for developing models for longitudinal #spatialtranscriptomics in the #retina? Join our team @hertie-ai.bsky.social at @unituebingen.bsky.social @ml4science.bsky.social!
jobs.medizin.uni-tuebingen.de/Job/5995/Pos...
jobs.medizin.uni-tuebingen.de/Job/5995/Pos...
March 14, 2025 at 2:44 PM
Are you looking for a #PhD or #Postdoc in #bioinformatics and #machinelearning for developing models for longitudinal #spatialtranscriptomics in the #retina? Join our team @hertie-ai.bsky.social at @unituebingen.bsky.social @ml4science.bsky.social!
jobs.medizin.uni-tuebingen.de/Job/5995/Pos...
jobs.medizin.uni-tuebingen.de/Job/5995/Pos...
We are looking for an HPC Compute Cluster administrator to contribute to our ML Cloud! jobs.medizin.uni-tuebingen.de/Job/3574/HPC...
November 8, 2023 at 9:13 AM
We are looking for an HPC Compute Cluster administrator to contribute to our ML Cloud! jobs.medizin.uni-tuebingen.de/Job/3574/HPC...