Biomedical image segmentation is critical for precise structure delineation and downstream analysis. Traditional methods often struggle with noisy data, while deep learning models such as U-Net have set new benchmarks in segmentation performance. nnU-Net further automates model configuration, making it adaptable across datasets without extensive tuning. However, it requires a substantial amount of annotated data for cross-validation, posing a challenge when only raw images but no labels are available. Large foundation models offer zero-shot generalizability, but may underperform on specific datasets with unique characteristics, limiting their direct use for analysis. This work addresses these bottlenecks by proposing a data-centric AI workflow that leverages active learning and pseudo-labeling to combine the strengths of traditional neural networks and large foundation models while minimizing human intervention. The pipeline starts by generating pseudo-labels from a foundation model, which are then used for nnU-Net's self-configuration. Subsequently, a representative core-set is selected for minimal manual annotation, enabling effective fine-tuning of the nnU-Net model. This approach significantly reduces the need for manual annotations while maintaining competitive performance, providing an accessible solution for biomedical researchers to apply state-of-the-art AI techniques in their segmentation tasks. The code is available at https://github.com/MMV-Lab/AL_BioMed_img_seg.

The challenge of labeling large example datasets for computer vision continues to limit the availability and scope of image repositories. This research provides a new method for automated data collection, curation, labeling, and iterative training with minimal human intervention for the case of overhead satellite imagery and object detection. The new operational scale effectively scanned an entire city (68 square miles) in grid search and yielded a prediction of car color from space observations. A partially trained yolov5 model served as an initial inference seed to output further, more refined model predictions in iterative cycles. Soft labeling here refers to accepting label noise as a potentially valuable augmentation to reduce overfitting and enhance generalized predictions to previously unseen test data. The approach takes advantage of a real-world instance where a cropped image of a car can automatically receive sub-type information as white or colorful from pixel values alone, thus completing an end-to-end pipeline without overdependence on human labor.

The future of software development will be model-driven, not code-driven. Now that my 4th book ("The Economics of Data, Analytics and Digital Transformation") is in the hands of my publisher, it's time to get back to work investigating and sharing new learnings. In this blog I'll take on the subject of Software 2.0. If the animal has four legs (except when it only has 3 legs due to an accident), and if the animal has short fur (except when it is a hair dog or a chihuahua with no fur), and if the animal has medium length ears (except when the dog is a bloodhound), and if the animal has a medium length legs (except when it's a bull dog), and if… Well, you get the point. In fact, it is probably impossible to distinguish a dog from other animals coding in if-then statements.

Like many ML practitioners, I started my ML journey with Kaggle competitions. But the comfortable setup of Kaggle, where you are handed largely clean data along with features and labels, could not be further from the reality of today's ML practitioner. In fact, the model building process itself is merely a small fraction of the work that needs to be done when developing an ML solution and deploying and maintaining it in production. It is useful to speak about ML production systems in terms of various degrees of maturity, where the least mature systems are one-off models, and the most mature systems run on autopilot, updating themselves with minimal human intervention. Here, I make a broad categorization of ML systems into four levels of increasing maturity, and discuss some of the challenges involved at each level. Disclaimer: given the choice of medium (a blog post, not a book chapter), this list will certainly be incomplete, and I didn't intend it to be.

The future of software development will be model-driven, not code-driven. Now that my 4th book ("The Economics of Data, Analytics and Digital Transformation") is in the hands of my publisher, it's time to get back to work investigating and sharing new learnings. In this blog I'll take on the subject of Software 2.0. If the animal has four legs (except when it only has 3 legs due to an accident), and if the animal has short fur (except when it is a hair dog or a chihuahua with no fur), and if the animal has medium length ears (except when the dog is a bloodhound), and if the animal has a medium length legs (except when it's a bull dog), and if… Well, you get the point. In fact, it is probably impossible to distinguish a dog from other animals coding in if-then statements.

While Strategy may not be dead, the importance of the traditional strategy function will be greatly reduced by the emergence of AI. In my previous blog "Strategy Series Part 1: "Creating a Data Strategy that Delivers Value," I asked: How does one develop data and AI strategies in a world of continuous change and transformation? Tesla provides an interesting poster child for that question: What is Tesla's business strategy? Is Tesla an automotive company or a transportation company or a logistics company, or what? At its heart, Tesla is in the data business.

The COVID-19 crisis has hammered home the importance for organizations to become more digital. And I suspect that most organizations are thinking that just means being able to support remote customer engagements and business operations. Organizations that are thriving during COVID-19 are those that have gone beyond just "digitalizing" their engagements and operations, but are actively leveraging granular customer, product and operational data to build analytic profiles (digital twins) around which they can optimize key business processes and uncover new revenue opportunities. For example, OpenTable is helping their customers avoid long lines at grocery stores. OpenTable provides a free restaurant reservations app for diners.

Uber and Lyft are probably the most visible examples of algorithms for managing people. These companies use various algorithms to efficiently find the closest drivers, the best routes, and the most effective methods of transportation for their customers. This is with minimal human intervention. Uber drivers don't ever really need to interact with their human bosses. Instead, their livelihoods are dictated by a model far off in a server somewhere telling them they need to go to 4th Street in rush hour to pick up Jamie who needs to go the airport.

Embark has just completed a coast-to-coast test drive of its self-driving semi-truck. The San Francisco-based startup is building self-driving software specifically designed for long-haul routes on interstate highways. The 2,400-mile ride lasted 44 hours over the course of 5 days, as there had to be a human pilot ready to takeover in case of a disengagement. For more videos, subscribe to Mashable Daily: http://on.mash.to/SubscribeNews Give us a follow: Facebook: https://www.facebook.com/mashable/

But what if your next employee didn't have these requirements. What if you could just show them how to do their job once and they turned around and worked around the clock with 100% accuracy? If you hate efficiency and saving money, stop now. You will only be very upset if you continue reading. But what if your next employee didn't have these requirements.