AI
Building robust AI models for the real world has a dirty secret: it’s rarely about the model architecture anymore. It’s almost always about the data. In the realm of audio, specifically event detection like identifying a dog’s cough versus a bark, data is noisy, scarce, and incredibly
AI
When we added new audio classes to our MelCNN pipeline (sneezing, coughing, anomalies), something unintuitive happened: accuracy on existing, well-performing classes like eating and drinking dropped - even though their data hadn’t changed. At first glance, this looked like a training issue. It wasn’t. What Actually Broke 1.
AI
Building high-quality audio datasets is a uniquely painful challenge. Unlike images, where you can glance at a grid of thumbnails and instantly spot a "bad" sample, audio requires time. You have to listen. Linear time is the enemy; you cannot "fast forward" through a dataset validation
AI
"We need at least a million labeled images to train this model." If you've ever heard this from a data scientist and felt your heart sink, you're not alone. The reality of building production ML systems is that you rarely have the luxury of
AI
The best engineering solutions aren't always the most complex ones. Sometimes, the smartest approach is knowing when to keep it simple—and when to add just enough redundancy to make it bulletproof. At Hoomanely, we build preventive pet healthcare technology that helps pet parents stay ahead of health
AI
How MelCNN learned “ghost sounds” from mislabeled silence Introduction Hallucination is usually discussed in the context of large language models—confident answers to questions that were never asked, or facts that never existed. But hallucination isn’t exclusive to text models. In audio systems, especially CNN‑based classifiers like MelCNN,
AI
Using large models to sanity-check edge detections Introduction Audio datasets are fragile. Unlike images or text, you cannot visually skim through thousands of audio samples and immediately know whether they belong to the right class. A barking dog, a metal clank, or background human speech can sound deceptively similar in
AI
The Calibration Problem Nobody Talks About When you ship a hardware product with sensors, you inherit a problem that pure software companies never face: the real world is messy. At Hoomanely, our smart bowl—Everbowl—uses load cells to measure food and water consumption and thermal sensors to track a
AI
Problem Modern pet‑health systems increasingly rely on thermal cameras to extract physiological signals such as eye temperature. The challenge is that thermal sensors are low‑resolution and almost never share the same optical center, field of view, or distortion model as the RGB camera doing the visual perception. In
AI
The Uncomfortable Truth About ML Projects Every machine learning project has a dirty secret: the model is usually not the problem. When a model underperforms, the instinct is to reach for a bigger architecture, more layers, or fancier training tricks. But in our experience building machine learning systems at Hoomanely,