# Convolutional-implementation-of-sliding-windows

**URL:** <https://community.deeplearning.ai/t/convolutional-implementation-of-sliding-windows/878359>\
**Category:** Convolutional Neural Networks\
**Tags:** week-module-3\
**Created:** [September 3, 2025, 6:02am UTC](https://community.deeplearning.ai/t/convolutional-implementation-of-sliding-windows/878359 "2025-09-03T06:02:48Z")\
**Posts on this page:** 8\
**Page:** 1

<div class="post-metadata">

**Author:** ![Mohamad\_Omar](https://avatars.discourse-cdn.com/v4/letter/m/ba8739/32.png) [@Mohamad\_Omar](https://community.deeplearning.ai/u/Mohamad_Omar)\
**Post date:** [September 3, 2025, 6:02am UTC](https://community.deeplearning.ai/t/convolutional-implementation-of-sliding-windows/878359/1 "2025-09-03T06:02:48Z")

</div>

please support on how Prof.Andrew converting FC to Conv, I have been watched the tutorial.

Just , I want the rule/Method in how he doing the converting ?

I see that he make conv by put 1\*1\*400 .400 is counting # of filters .

---

<div class="post-metadata">

**Author:** ![gent.spah](https://sea1.discourse-cdn.com/dlai/user_avatar/community.deeplearning.ai/gent.spah/32/136423_2.png) [@gent.spah](https://community.deeplearning.ai/u/gent.spah)\
**Post date:** [September 3, 2025, 6:07am UTC](https://community.deeplearning.ai/t/convolutional-implementation-of-sliding-windows/878359/2 "2025-09-03T06:07:23Z")

</div>

Which video is this referring to?

---

<div class="post-metadata">

**Author:** ![Mohamad\_Omar](https://avatars.discourse-cdn.com/v4/letter/m/ba8739/32.png) [@Mohamad\_Omar](https://community.deeplearning.ai/u/Mohamad_Omar)\
**Post date:** [September 3, 2025, 6:08am UTC](https://community.deeplearning.ai/t/convolutional-implementation-of-sliding-windows/878359/3 "2025-09-03T06:08:42Z")

</div>

PFB:

 ![image](https://global.discourse-cdn.com/dlai/original/3X/8/b/8b008dd950dacbfa47c1e960de359700b7e59838.jpeg)

---

<div class="post-metadata">

**Author:** ![gent.spah](https://sea1.discourse-cdn.com/dlai/user_avatar/community.deeplearning.ai/gent.spah/32/136423_2.png) [@gent.spah](https://community.deeplearning.ai/u/gent.spah)\
**Post date:** [September 3, 2025, 6:15am UTC](https://community.deeplearning.ai/t/convolutional-implementation-of-sliding-windows/878359/4 "2025-09-03T06:15:23Z")

</div>

Here I used chatgpt to make the calculation of convolutions for the first 2 layers:

 ![Screenshot 2025-09-03 at 10.16.17 AM](https://global.discourse-cdn.com/dlai/original/3X/7/a/7aa647f09536466a9e187e89cbfe3a4df4733797.png)

 ![Screenshot 2025-09-03 at 10.16.57 AM](https://global.discourse-cdn.com/dlai/original/3X/6/a/6ae21efd06b2da3df632d9cbf50c01a997e65a3c.png)

the same thing you can do when you go along with maxpoling and then flatten the entire 3D matrix into a vector!

---

<div class="post-metadata">

**Author:** ![Mohamad\_Omar](https://avatars.discourse-cdn.com/v4/letter/m/ba8739/32.png) [@Mohamad\_Omar](https://community.deeplearning.ai/u/Mohamad_Omar)\
**Post date:** [September 3, 2025, 6:35am UTC](https://community.deeplearning.ai/t/convolutional-implementation-of-sliding-windows/878359/5 "2025-09-03T06:35:02Z")

</div>

what you shared is conv steps for the first 2 layers but how he convert FC to Conv in final , is it same as you did ?i just shared a picture to show you the vedio and not where I’m fail to understand .wait i will share my problem in clear pic

---

<div class="post-metadata">

**Author:** ![Mohamad\_Omar](https://avatars.discourse-cdn.com/v4/letter/m/ba8739/32.png) [@Mohamad\_Omar](https://community.deeplearning.ai/u/Mohamad_Omar)\
**Post date:** [September 3, 2025, 6:52am UTC](https://community.deeplearning.ai/t/convolutional-implementation-of-sliding-windows/878359/6 "2025-09-03T06:52:30Z")

</div>

Please check the Highlighted Box in red , here i lost the focus and image start blur for me.

how he moved from **1 by 1 by 400** to **1 by 1 by 4**

 ![image](https://global.discourse-cdn.com/dlai/original/3X/7/8/788ac215e0b27672fd616cdffd0ba7b28ad7850f.jpeg)

---

<div class="post-metadata">

**Author:** ![Mohamad\_Omar](https://avatars.discourse-cdn.com/v4/letter/m/ba8739/32.png) [@Mohamad\_Omar](https://community.deeplearning.ai/u/Mohamad_Omar)\
**Post date:** [September 3, 2025, 7:15am UTC](https://community.deeplearning.ai/t/convolutional-implementation-of-sliding-windows/878359/7 "2025-09-03T07:15:36Z")

</div>

now i know how we get 400 , it’s by multiply 5\* _5_ \*16. i get support from chatgpt too

---

<div class="post-metadata">

**Author:** ![gent.spah](https://sea1.discourse-cdn.com/dlai/user_avatar/community.deeplearning.ai/gent.spah/32/136423_2.png) [@gent.spah](https://community.deeplearning.ai/u/gent.spah)\
**Post date:** [September 3, 2025, 8:55am UTC](https://community.deeplearning.ai/t/convolutional-implementation-of-sliding-windows/878359/8 "2025-09-03T08:55:42Z")

</div>

Its good to use for learning purposes too!