Show HN: Springus – Outfit recommendations from your real wardrobe using AI

Hey HN,

Since starting to work from home, I noticed my motivation to get dressed in the morning tanked. I’d default to the same sweatpants, which started affecting my mood and productivity. I wanted something to nudge me to dress better—without making it a chore.

That’s why I built Springus, a wardrobe companion for iOS. Instead of manually cataloguing every item, Springus uses a multi-class segmentation model to build your digital closet from fit pix. The recommendation system then suggests outfits from clothing you actually own, aiming to reduce decision fatigue and help you find combinations you might not have considered.

The hardest part was making the segmentation work reliably with real-world photos — messy backgrounds, bad lighting, and all. I ended up training a custom model on hundreds of my own fit pics and some of friends, iterating until it was good enough to share.

I’ve been using Springus every day for the last 2 months. It’s free, and there’s no catch — I plan to monetize later by recommending clothes that fit your style, but right now, it’s just a passion project I wanted to share.

If you’re interested, I’d love feedback — especially on the segmentation accuracy and the outfit recommendations. What would make this genuinely useful for you?