跳到主要内容
Learn Motion
探索课程产品能力价格

下载 Learn Motion

用手机扫描二维码,下载 App。

App Store

扫码从 App Store 下载 Learn Motion打开商店页面

Google Play

扫码从 Google Play 下载 Learn Motion打开商店页面
登录
Learn Motion

How AI Sees: A Conceptual Guide to Image and Video Understanding

1What a Machine Actually Sees2From Pixels to Patterns: Features3How Deep Networks Learn to See4Beyond Labels: Locating and Describing What Is Seen5Adding Time: Understanding Video6How These Systems Learn and How We Judge Them
From Pixels to Patterns: Features

Why Hand-Designed Features Break Down

4 / 4
Take the building example seriously. The detector was tuned on daytime photos, where the building's edges are strong and easy to find. At night, most of those edges shrink below the threshold and disappear, while a few lit windows create bright gradients that were never the structure you cared about. Nothing about the detector changed — the same fixed rule was applied to numbers that now mean something different. That is the real limitation. It is not that these features are wrong, it is that they are decided in advance, so every new condition needs new engineering, and fixing one case often breaks another.
0:00 / 0:00

Lighting

A gradient measures change in recorded intensity, not change in the physical world. Dim the scene and every change shrinks, so weak but real edges drop below the threshold. Brighten it unevenly and shadows produce strong gradients that mark nothing real. The same object, the same detector, two different results.

Pose and scale

A corner viewed head-on can appear as a gentle curve from an oblique angle, so the two-direction change disappears and the keypoint is lost. Scale causes a parallel failure: a detector tuned to change over a few pixels finds nothing when the object fills the frame, and a detector tuned to large structures misses fine detail.

The same building, two photographs

A detector tuned on daytime street photos of a building finds plenty of strong corners and edges. Photograph the same building at night, lit only by a few windows, and most of those edges vanish below threshold while the window frames produce bright gradients that were never the structure you cared about. The detector has not learned anything; it is applying the same fixed rule to numbers that now mean something different.

The limitation is not that hand-designed features are wrong, but that they are fixed. Because the measurements are chosen in advance, every new condition requires new engineering, and fixes for one condition often degrade another. That is the gap that learned features were meant to close.

上一页4 / 4下一页

Learn Motion

生成课程,认真学会。

运营方:Wuhan Daoyin Technology Co., Ltd.

联系我们: [email protected]
隐私政策用户服务协议

© 2026 Learn Motion