hello

@pikacodes

196.7K followers · usually 8.1K views a post

Views

26.9K

Likes

1.7K

Comments

16

Shares

40

SoundTikTok sound · original sound - pikacodes

Promoted product

Warp

Top 20 LTL carrier rebuilding freight from the ground up. Pooled freight across shippers. Prices drop as the network grows. The network belongs to the shippers who use it.

3 videos2 creators

Why it worked

The video effectively showcases a practical application of an AI tool (Warp) in solving a common developer problem (data labeling for machine learning). The creator's clear explanation and demonstration of how Warp streamlines a tedious task make it relatable and valuable to other developers.

Summary

The creator demonstrates how they used Warp, an AI-powered terminal, to improve their ASL detection app. Warp helped them write code to save misclassified frames and relabel them, streamlining the data collection process for retraining their model.

Structure

1Introduce ASL detection app
2Show manual fixing of errors
3Introduce Warp as a solution
4

Original caption

🤘 improving my ASL detection app with @warp.dev :] #ad #programming #ai #coding #machinelearning

More from @pikacodes

Product placement

Warp: it acts as a modern terminal with a built-in AI mode that helps the creator code faster. This is the promoted product. Warp is used to generate code and scripts. Removing it would change what happens in the video. OpenCV: it appears as a library used in the code. Removing it would not change what happens in the video.

Call to action

check out Warp and let me know what you think :)

On-screen text

hello i love you i love you this sign should be "bullsh*t" 😭 manually fixing and adding new signs the hard way let's improve the model with a little help from my friend it's a modern terminal with a built-in AI mode to help me save an opencv frame and string label with a key press the goal is to reuse these images later and relabel them correctly write a function that saves the current frame whenever and automatically stores it in a folder with the label name and a timestamp now I can capture misclassified frames actually assign them to their correct label so again I asked Warp to generate a relabeling script so I can immediately correct the mislabeled image it helped me build a simple CLI tool that helps me create new labels or assign the images to an existing one then it moves the file to my data folder so it's ready for when I retrain my model what I love about Warp these small but really repetitive tasks easier to do is ask in plain English there's nothing a dev loves more than automation AUTOMATION ROCKS after this retraining and ta-dah! check out Warp and let me know what you think :)

Transcript

I coded detection app works really well for the most part manually fixing and adding new signs the hard way let's improve the model with a little help from my friend it's a modern terminal with a built-in AI mode to help me save an opencv frame and string label with a key press the goal is to reuse these images later and relabel them correctly write a function that saves the current frame whenever and automatically stores it in a folder with the label name and a timestamp now I can capture misclassified frames actually assign them to their correct label so again I asked Warp to generate a relabeling script so I can immediately correct the mislabeled image it helped me build a simple CLI tool that helps me create new labels or assign the images to an existing one then it moves the file to my data folder so it's ready for when I retrain my model what I love about Warp these small but really repetitive tasks easier to do is ask in plain English there's nothing a dev loves more than automation AUTOMATION ROCKS after this retraining and ta-dah! it works check out Warp and let me know what you think :)