How to achieve "depth" with a 3d css transform

Here is a hacky idea using multiple background to simulate such effect. The trick is to add 2 semi-transparent gradients to create the shadow effect then 2 other gradient to cut a small part of the corner to obtain the 3D shape.

The result may not be perfect for all the images:

.wrapper {
  display:inline-block;
  perspective:1000px;
}
.box {
  margin: 50px;
  width:200px;
  height:200px;
  transform: rotate3d(360, 120, -90, 60deg) rotateZ(-30deg);
  background:
  linear-gradient(to bottom right,transparent 49%,#fff 52%) bottom right/14px 10px,
  linear-gradient(to top left,transparent 49%,#fff 52%) top left /10px 14px,
  
  linear-gradient(rgba(0,0,0,0.5),rgba(0,0,0,0.5)) 0 0px/10px 100%,   
  linear-gradient(rgba(0,0,0,0.5),rgba(0,0,0,0.5)) 100% 100%/calc(100% - 10px) 10px,   
  url(https://picsum.photos/id/1061/1000/800) center/cover;
  background-repeat:no-repeat;
}
<div class="wrapper" >
<div class="box" >
</div>
</div>

With your image you can have a specific gradient like below:

body {
  background:#ccc;
  
}
.wrapper {
  display:inline-block;
  perspective:1000px;
}
.box {
  margin: 50px;
  width:200px;
  height:250px;
  transform: rotate3d(360, 120, -90, 60deg) rotateZ(-30deg);
  background:
  linear-gradient(to bottom right,transparent 49%,#ccc 52%) bottom right/16px 10px,
  linear-gradient(to top left,transparent 49%,#ccc 52%) top left /10px 12px,
  
  linear-gradient(#efefef,#efefef) 100% 100%/calc(100% - 10px) 10px, 
  linear-gradient(-226deg,#222428 13px,#ff4946 13px,#ff4946 77px,#592D30 77px,#592D30 100px,#222428 100px,#222428 108px,#efefef 108px,#efefef 161px) 0 0px/10px 100%,      
  
  url(https://i.imgur.com/foDEYpB.png) center/cover;
  background-repeat:no-repeat;
}
<div class="wrapper">
  <div class="box">
  </div>
</div>

Try adding three type of images to make 3D effects. Use transform property with rotation for images to get the desired result.

Answer reference here.

.perspective {
    position: relative;
    width: 400px;
    height: 500px;
    transform-style: preserve-3d;
    transition: all 500ms ease-in;
    transform: rotateY(20deg) rotateX(60deg) rotateZ(-10deg);
    transform: rotateY(15deg) rotateX(50deg) rotateZ(-15deg);
    box-shadow: -40px 80px 80px -10px rgba(0, 0, 0, 0.7);
    cursor: pointer;
    margin-right: 30px;
    display: inline-block;
    margin-left: 30%;
}

.perspective img {
    position: absolute;
    top: 0px;
    left: 0px;
    width: 400px;
    height: 500px;
    transform: translateZ(16px);
}

.bottom,
.left {
    position: absolute;
    width: 400px;
    height: 500px;
    display: block;
    transition: all 1s linear;
    overflow: hidden;
    border-radius: 3px;
    transform: translateZ(16px);
    filter: brightness(80%)
}

.left {
    transform: rotateY(270deg) translateX(-1px);
    transform-origin: center left;
    width: 18px;
}

.bottom {
    transform: rotateX(90deg) translateY(15px) translateZ(-480px);
    transform-origin: bottom center;
    height: 18px;
}

.bottom img {
    transform: rotateX(180deg);
    width: 100%;
    height: 500px;
    left: 0px;
}
<div class="perspective">
    <img src="https://i.imgur.com/foDEYpB.png">
    <div class="bottom"><img src="https://i.imgur.com/foDEYpB.png"></div>
    <div class="left"><img src="https://i.imgur.com/foDEYpB.png"></div>
</div>