Urgent.News

What's breaking now, across thousands of outlets.

Tech

TrailWing: An Offline Bird Call Identifier That Works Where Your Signal Doesn't

What I Built Problem statement: "A bird call identifier that works on the trail with no signal." TrailWing is an open-source, offline-first birding companion. You pick a trail or park before you leave home, and the app downloads a small "trail pack": the bird species likely to be around this week, plus the local weather forecast. After that, you put your phone in your pocket and walk. When you…

TrailWing is an offline-first birding companion app that works on trails without needing cell service. To use it, participants select a trail or park ahead of time, and the app downloads a "trail pack" containing the bird species likely to be found in that location during a specific week. After downloading, the user simply walks along the trail, presses a button when they hear a bird, holds the phone up for about ten seconds, and then puts the phone away again.

An on-device model compares the audio clip to the species listed in the trail pack and provides a brief response, such as "Probably a Black-capped Chickadee, 82% confident," along with one line on how to confirm the sighting visually. The app is intended for beginners and intermediate birders, hikers, and families who want to learn about the birds in their surroundings without relying on cellular connectivity or sharing their location with a remote server.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in Tech

I tried 13 OCR preprocessing tricks on screenshots. None helped

Almost every OCR tip I found said the same thing. Before recognizing a screenshot, clean it up: go grayscale, binarize it, run a denoise filter, maybe upscale it 2x.

  • Thirteen OCR preprocessing techniques tested on screenshots, none improved results
  • Grayscale conversion worsened CER from 0.1% to 9.7% in Professional OCR tier
  • Vertical text recognition not helped by any preprocessing techniques

Launch day is the worst day to judge a product

Most launches I've watched, my own included, have the same shape. Day one is a screenshot, a one-liner and a link. And the product on that day is the least it will ever be: half the settings are…

  • First day is worst for judging product quality
  • Features incomplete, onboarding vague
  • By third week, product significantly improves

Measure ink and lighting before you binarize an image for OCR

My team is looking at client-side OCR for an internal tool where people attach screenshots and photos of printed notices.

  • Measure ink darkness and paper brightness variation before thresholding
  • Fixed threshold discards light text and blackens uneven lighting areas
  • Measure ink density, paper brightness percentile, and Otsu threshold value

More from Saturday 10 October →