Our work on validating language-vision learning by collecting large-scale text-image LAION-5B dataset and training various scale CLIP models, evaluating those on different downstream tasks like zero-shot classification, won Outstanding Paper Award at NeurIPS 2022. Work was performed at Scalable Learning & Multi-Purpose AI Lab (Mehdi Cherti & Jenia Jitsev) in collaboration with LAION, UC Berkeley, TU Darmstadt, TU München, University of Washington, Allen AI Institute.<br>
Eric Upschulte won a **Winner Finalist Award at NeurIPS 2022's Cell segmentation challenge**.
In this Journal Club we will discuss the according [paper](https://openreview.net/forum?id=YtgRjBw-7GJ). Eric will give a **short intro** and is happy to answer your questions about his solution and what it's like to take part in a NeurIPS competition.
We present LAION-5B, an open, publically available dataset of 5.8B image-text pairs and validate it by reproducing results of training state-of-the-art CLIP models of different scale.<br>
The paper present a simple framework for cell segmentation, based on uncertainty-aware Contour Proposal Networks (CPNs). It is designed to provide high segmentation accuracy while remaining computationally efficient, which makes it an ideal solution for high throughput microscopy applications. Each predicted cell is provided with four uncertainty estimations that give information about the localization accuracy of the detected cell boundaries. Such additional insights are valuable for downstream single-cell analysis in microscopy image-based biology and biomedical research.
In the context of the NeurIPS 22 Cell Segmentation Challenge, the proposed solution is shown to generalize well in a multi-modality setting, while respecting domain-specific requirements such as focusing on specific cell types.<br>
### 23 January 2023: LAION-5B: An open large-scale dataset for training next generation image-text models - Outstanding Paper Award at NeurIPS 2022
Jenia Jitsev, Mehdi Cherti (JSC)
Our work on validating language-vision learning by collecting large-scale text-image LAION-5B dataset and training various scale CLIP models, evaluating those on different downstream tasks like zero-shot classification, won Outstanding Paper Award at NeurIPS 2022. Work was performed at Scalable Learning & Multi-Purpose AI Lab (Mehdi Cherti & Jenia Jitsev) in collaboration with LAION, UC Berkeley, TU Darmstadt, TU München, University of Washington, Allen AI Institute.<br>
We present LAION-5B, an open, publically available dataset of 5.8B image-text pairs and validate it by reproducing results of training state-of-the-art CLIP models of different scale.<br>
