Papers by Raphael Schumann
Discrete Optimization for Unsupervised Sentence Summarization with Word-Level Extraction (2020.acl-main)
Copied to clipboard
| Challenge: | Sentence summarization systems that use latent space to reconstruct the source sentence are unwillingly exploited. |
| Approach: | They propose a method that uses language modeling and semantic similarity metrics to find a high-scoring summary. |
| Outcome: | The proposed method achieves state-of-the-art for unsupervised sentence summarization according to ROUGE scores. |
Scalable Wide and Deep Learning for Computer Assisted Coding (N18-3)
Copied to clipboard
Marilisa Amoia, Frank Diehl, Jesus Gimenez, Joel Pinto, Raphael Schumann, Fabian Stemmer, Paul Vozila, Yi Zhang
| Challenge: | In recent years the use of electronic medical records has accelerated resulting in large volumes of medical data when a patient visits a healthcare facility. |
| Approach: | They propose to use convolutional neural networks and logistic regression to build a machine learning based system for predicting ICD-10 codes from electronic medical records. |
| Outcome: | The proposed system can predict ICD-10 codes from electronic medical records using convolutional neural networks and logistic regression models. |
Backward Compatibility During Data Updates by Weight Interpolation (2024.eacl-long)
Copied to clipboard
| Challenge: | Retraining a model with a larger amount of training data introduces negative flips . retraining the model with the updated data introduce negative flipping . |
| Approach: | They propose a backward compatible weight interpolation method to improve model predictions without regression bugs. |
| Outcome: | The proposed method reduces negative flips without sacrificing accuracy . it is straight forward to implement and does not increase inference cost. |
Generating Landmark Navigation Instructions from Maps as a Graph-to-Text Problem (2021.acl-long)
Copied to clipboard
| Challenge: | Current navigation services based on turns and distances of named streets . humans use efficient mode of navigation based around visible and salient landmarks . |
| Approach: | They propose a neural model that takes OpenStreetMap representations as input and learns to generate navigation instructions that contain salient landmarks from human natural language instructions. |
| Outcome: | The proposed model can generate navigation instructions that contain salient landmarks from openStreetMap . it is based on a dataset of 7,672 instances verified by human navigation in Street View . |
Analyzing Generalization of Vision and Language Navigation to Unseen Outdoor Areas (2022.acl-long)
Copied to clipboard
| Challenge: | Recent work on visual-grounded navigation has focused on indoor scenarios with sharp drops in performance when testing on unseen data. |
| Approach: | They focus on visual agent navigation in outdoor scenarios with panorama images . they find that most gain in outdoor VLN on unseen data is due to specific features . |
| Outcome: | The results show a bias to specifics of graph representations of urban environments, demanding that VLN tasks grow in scale and diversity of geographical environments. |