<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Nous Technology Limited</title>
    <description>The latest articles on DEV Community by Nous Technology Limited (@nousrun).</description>
    <link>https://dev.to/nousrun</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4013899%2Fdbb0f3fb-79c8-4a67-8ed0-d4634eee4182.png</url>
      <title>DEV Community: Nous Technology Limited</title>
      <link>https://dev.to/nousrun</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/nousrun"/>
    <language>en</language>
    <item>
      <title>What Ball Tracking Misses in Tennis Video Analysis</title>
      <dc:creator>Nous Technology Limited</dc:creator>
      <pubDate>Fri, 03 Jul 2026 17:13:23 +0000</pubDate>
      <link>https://dev.to/nousrun/what-ball-tracking-misses-in-tennis-video-analysis-2m1k</link>
      <guid>https://dev.to/nousrun/what-ball-tracking-misses-in-tennis-video-analysis-2m1k</guid>
      <description>&lt;p&gt;Most tennis video analysis starts with the ball.&lt;/p&gt;

&lt;p&gt;Where did it land?&lt;br&gt;&lt;br&gt;
How fast did it move?&lt;br&gt;&lt;br&gt;
Was the shot in or out?&lt;br&gt;&lt;br&gt;
What was the trajectory?&lt;/p&gt;

&lt;p&gt;Those are useful questions. But they are not the whole story.&lt;/p&gt;

&lt;p&gt;In tennis, a player can hit a clean-looking shot and still be late for the next ball. A rally can start breaking down before the ball leaves the racket. The visible result may be the shot, but the cause often sits inside the athlete’s movement: timing, balance, recovery, and readiness.&lt;/p&gt;

&lt;p&gt;That is the layer we are exploring with SpatialForm.&lt;/p&gt;

&lt;p&gt;We are building SpatialForm as an early AI sports product at NOUS TECHNOLOGY LIMITED. It focuses on turning ordinary phone sports video into &lt;strong&gt;Performance Form&lt;/strong&gt; — a visible movement layer for reviewing timing, balance, recovery, and next-ball readiness.&lt;/p&gt;

&lt;p&gt;Official site: &lt;a href="https://www.nous.run/" rel="noopener noreferrer"&gt;SpatialForm official site&lt;/a&gt;&lt;/p&gt;




&lt;h2&gt;
  
  
  Ball tracking explains the result
&lt;/h2&gt;

&lt;p&gt;Ball tracking is valuable because it turns the most obvious part of tennis into data. It can help answer questions like where the ball landed, how fast it moved, and what happened after contact.&lt;/p&gt;

&lt;p&gt;For players, this is useful. For coaches, it creates a clearer record of the rally.&lt;/p&gt;

&lt;p&gt;But the ball is often the receipt, not the transaction.&lt;/p&gt;

&lt;p&gt;By the time the ball has landed, the movement decision has already happened. A player may lose the rally because the split step was late. Because the first step went in the wrong direction. Because the body was off-balance at contact. Because recovery was slow.&lt;/p&gt;

&lt;p&gt;None of those problems are fully explained by the ball alone.&lt;/p&gt;




&lt;h2&gt;
  
  
  Tennis often breaks down before contact
&lt;/h2&gt;

&lt;p&gt;Many tennis mistakes look like shot problems. A forehand goes long. A backhand lands short. A clean shot does not lead to control of the point.&lt;/p&gt;

&lt;p&gt;But if you look earlier in the sequence, the issue may not be the swing itself. It may be:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;the player loaded too late&lt;/li&gt;
&lt;li&gt;the first step was delayed&lt;/li&gt;
&lt;li&gt;the contact point was reached under poor balance&lt;/li&gt;
&lt;li&gt;the body was still moving away from the next ball&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;This is why normal replay is limited. A replay can show what happened, but it does not always make the movement structure obvious. The important frame is often not the contact frame. It may be the frame before contact. Or three frames before contact.&lt;/p&gt;

&lt;p&gt;That is where the athlete layer starts to matter.&lt;/p&gt;




&lt;h2&gt;
  
  
  The athlete layer behind every shot
&lt;/h2&gt;

&lt;p&gt;When we say “athlete layer,” we mean the visible movement structure behind the shot. Not medical diagnosis. Not lab-grade biomechanics. Not replacing a coach.&lt;/p&gt;

&lt;p&gt;We mean practical movement signals that can be reviewed from video:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;timing &amp;amp; loading&lt;/li&gt;
&lt;li&gt;balance &amp;amp; transfer&lt;/li&gt;
&lt;li&gt;contact stability&lt;/li&gt;
&lt;li&gt;recovery &amp;amp; court re-entry&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;This is the gap between shot analysis and Performance Form. Shot tracking explains the ball. Performance Form explains whether the athlete was ready for the next action.&lt;/p&gt;

&lt;p&gt;More on Performance Form: &lt;a href="https://www.nous.run/PerformanceForm/" rel="noopener noreferrer"&gt;Performance Form&lt;/a&gt;&lt;/p&gt;




&lt;h2&gt;
  
  
  Why phone video is a tough computer vision problem
&lt;/h2&gt;

&lt;p&gt;Using ordinary phone video sounds simple, but it creates several hard engineering problems. Phone videos are not recorded in controlled lab conditions.&lt;/p&gt;

&lt;p&gt;For a computer vision system, they bring:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;unstable camera angles &amp;amp; motion blur&lt;/li&gt;
&lt;li&gt;changing lighting &amp;amp; partial occlusion&lt;/li&gt;
&lt;li&gt;racket and limb overlap&lt;/li&gt;
&lt;li&gt;inconsistent player distance from the camera&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The harder challenge is turning noisy visual information into a useful movement review. Standard pose estimation models can struggle here because a practical tennis movement analysis system needs to understand &lt;em&gt;time&lt;/em&gt;, not just shape.&lt;/p&gt;

&lt;p&gt;It needs to ask: what happened before contact? Did the player recover? Was the movement sequence repeatable? A single pose frame is not enough. The value comes from the sequence.&lt;/p&gt;




&lt;h2&gt;
  
  
  From phone sports video to Performance Form
&lt;/h2&gt;

&lt;p&gt;SpatialForm is built around the idea that ordinary sports video can reveal more than a replay. The goal is to transform phone video into a reviewable movement signal.&lt;/p&gt;

&lt;p&gt;At a high level, that means looking beyond a single shot result and focusing on the athlete’s movement sequence:&lt;/p&gt;

&lt;p&gt;&lt;code&gt;load → prepare → contact → recover → reset&lt;/code&gt;&lt;/p&gt;

&lt;p&gt;For tennis, the important question is not only: &lt;em&gt;Where did the ball go?&lt;/em&gt;&lt;br&gt;&lt;br&gt;
It is also: &lt;em&gt;Was the athlete ready for the next action?&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;Starting with tennis, the goal is to make these movement signals easier to see from the videos players and coaches already record.&lt;/p&gt;

&lt;p&gt;Tennis Swing Analysis page: &lt;a href="https://www.nous.run/TennisSwingAnalysis/" rel="noopener noreferrer"&gt;Tennis Swing Analysis&lt;/a&gt;&lt;/p&gt;




&lt;h2&gt;
  
  
  Why this matters for players and coaches
&lt;/h2&gt;

&lt;p&gt;Most amateur players already record video, but it is deeply underused. A player may watch a clip and say: &lt;em&gt;"I was late. My balance was off."&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;The problem is that these observations are often vague. They are felt, but not clearly seen. If movement signals can be made more visible, video review becomes actionable. A coach can point to a moment before contact. A player can compare recovery between two shots.&lt;/p&gt;

&lt;p&gt;This does not remove the human coach. It gives the player and coach a clearer visual layer to discuss.&lt;/p&gt;




&lt;h2&gt;
  
  
  What we are building
&lt;/h2&gt;

&lt;p&gt;The product direction is simple:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Phone Sports Video → Performance Form → Personal Athletic Intelligence&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;Instead of only asking what happened to the ball, SpatialForm asks what happened to the athlete behind the shot. The long-term goal is to help players understand movement patterns that are hard to see in normal replay.&lt;/p&gt;

&lt;p&gt;This is still early. We are not claiming that a phone replaces a professional biomechanics lab. The goal is more practical: Can ordinary sports video become a better movement review tool?&lt;/p&gt;

&lt;p&gt;That is the question we are exploring.&lt;/p&gt;

&lt;p&gt;Official site: &lt;a href="https://www.nous.run/" rel="noopener noreferrer"&gt;SpatialForm official site&lt;/a&gt;&lt;/p&gt;




&lt;h2&gt;
  
  
  Closing thought
&lt;/h2&gt;

&lt;p&gt;Ball tracking is useful. But in tennis, the ball is only one part of the story.&lt;/p&gt;

&lt;p&gt;The next layer of sports video analysis is not just where the ball went. It is whether the athlete was ready for the next action.&lt;/p&gt;

&lt;p&gt;That is the layer SpatialForm is working to make visible from ordinary phone sports video.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>computervision</category>
      <category>sports</category>
      <category>startup</category>
    </item>
  </channel>
</rss>
