January 2021 Summaries

Clarifai Release 7.1 introduces significant enhancements and new features across its platform, catering to developers with tools for text search, data management, object tracking, and deep training. Key updates include improvements in Explorer for easier label management and hierarchical concept modeling, alongside new text search capabilities that support custom concept searches and multilingual text moderation in over 52 languages. The release also offers deep training options for Professional Plan users, allowing detailed control over model parameters, and introduces video support with object and people tracking functionalities. Additionally, pre-validation of names for models and apps ensures a smooth user experience within the platform.

The Neural Information Processing Systems (NeurIPS) annual meeting is a prestigious AI conference where researchers present significant advancements in neural information processing. Notable papers from NeurIPS 2020 include one on RescaleNet, which proposes a new residual operation as an alternative to BatchNorm for improving model performance while addressing the vanishing/exploding variable problem in deep neural networks. Another paper introduces a method for uncertainty-aware self-training in few-shot text classification, using Monte Carlo Dropout for uncertainty estimation and hard example selection to enhance model training. Additionally, a paper on crowd counting redefines the task as a distribution matching problem, using a density map approach that combines various loss functions to outperform existing methods, particularly on large-scale datasets.