Does My Head Look Big In This

The App That Was Basically Google's First Serious Try-On Experiment

There used to be an app called Does My Head Look Big In This that let you point your phone camera at your face and overlay sunglasses, hats, and other accessories in real time. It was originally a Google Labs project around 2010, then picked back up years later as they were building the Glass prototype. The concept was straightforward — use facial landmark detection to track your face and anchor virtual objects to the right spots. It worked well enough for a proof of concept and made people laugh at tech demos everywhere. The core pipeline is pretty simple once you strip away the marketing. The app takes a video feed from your front-facing camera, runs a face detection model to find key landmarks like the bridge of your nose and the corners of your eyes, then renders a 3D model or image of whatever accessory you selected on top of those points. As you move your head, it applies perspective transforms so the glasses don't float off your face when you turn. The whole thing runs locally on the device, which is why it's fast and doesn't need a server connection after the initial load. I remember building a similar system a few years ago for a client who wanted virtual hat try-ons for an e-commerce brand. We used ARKit for the face tracking because it already had solid head mesh detection built in. The Google app did something comparable but with its own custom pipeline. The difference between a decent implementation and a terrible one usually comes down to one thing: how well the system handles poor lighting or extreme angles. Most apps that get downloaded once and never opened again fail at that checkpoint.

Where You Can Find It Now

The original Does My Head Look Big In This app from Google was eventually discontinued or folded into other projects. If you search the Apple App Store or Google Play Store, you'll find various clones and spin-offs that use the same name or similar functionality. Some are legitimate AR try-on tools, some are just slapped together with generic code and contain ads that make them unusable. The Google version specifically is no longer available for new downloads, though if you have it cached on your device or a backup, it may still run on older iOS versions. For people who just want the functionality without hunting through questionable app store listings, there are a handful of alternatives that do the same job better. Apple's Scene Kit-based sample projects from WWDC included a glasses try-on demo that you can run on your own device. Android has ARCore with its face effects API that does this natively. The open-source community also has several implementations using MediaPipe Face Mesh or TensorFlow Lite models that you can compile and run yourself if you're comfortable with that process.

Common Pitfalls People Run Into

Most users who try these apps and get frustrated share the same complaint: the virtual glasses slide off their face or don't sit right when they tilt their head. This isn't usually a bug in the app. It's a limitation of the underlying face tracking technology when dealing with certain face shapes, glasses wearers who already have frames in the camera view, or people with facial hair that obscures the landmark points the algorithm relies on. I once spent three days debugging why a client's hat overlay kept drifting to the left on about forty percent of test subjects. Turned out the training data the face detection model was built on was heavily skewed toward center-facing, symmetrical faces with no obstructions. The model simply couldn't reliably detect landmarks when someone had a wide-brimmed hat casting shadows across their forehead. Another issue that nobody talks about is the refresh rate. These apps need to maintain at least thirty frames per second for the overlay to feel anchored to your face. Anything lower and the brain registers the delay between your movement and the visual feedback, which makes everything look floaty and fake. On older devices running the original version of Does My Head Look Big In This, you'd often see frame drops during complex movements simply because the face detection model was too heavy for the GPU available at the time.

Does My Head Look Big In This for Actual Shopping Use

If you're asking whether this kind of technology is actually useful for buying sunglasses or hats online, the honest answer is that it's gotten significantly better but still has gaps. The main problem is scale and position accuracy. A virtual sunglasses model might look perfect in the demo because it was crafted on a mean face shape with ideal proportions. When you apply it to your actual face, the frame width might be off by a centimeter or two, which in the real world is the difference between glasses that fit and glasses that pinch. The AR overlay can't tell you about weight distribution, temple arm length, or how the frames will feel after wearing them for four hours. Those details still matter. That said, for a quick sanity check before buying, it works fine. I've used it to rule out styles that completely clash with my face shape before spending money on something I'd regret. The technology has improved enough over the last few years that the basic use case — trying on a handful of options and narrowing your choices — is genuinely helpful. It's not a replacement for actually trying the product on, but it's better than nothing and saves you from ordering five pairs and returning four of them. The apps that are worth your time tend to be the ones that invest in actual 3D models rather than flat image overlays. A flat PNG of sunglasses placed over your face will always look like a sticker. A proper 3D mesh that responds to lighting and perspective changes gives you a much more realistic preview. If the app you're using looks like those old Snapchat filters from 2016, it's probably not going to give you reliable information for a purchase decision.