Cafe, Analytic Scape (still), 2025, AI system, camera, video, 10 min, various installations.

Analytic Scape is a series of scape videos reconstructed through the eye of AI. After capturing various urban and natural scenes, the footage was analyzed using twenty distinct computer vision algorithms*. The elements of the world are classified and quantified by AI. The original image disappears, and only the analyzed data remains. This data manifests as labels, probability scores, outlines, colors, and directional lines. Human faces are marked with feature points and bounding boxes, objects are tagged with names, and movement is traced with directional vectors. As these layers intertwine and overlap, they form a scape of AI perception.

• Analytic Scape (documentation video), 2025.

• 

  Sea, Analytic Scape (still), 2025, AI system, camera, video, 10 min, various installations.

• 

  Subway, Analytic Scape (still), 2025, AI system, camera, video, 10 min, various installations.

• 

  Bird, Analytic Scape (still), 2025, AI system, camera, video, 10 min, various installations.

• 

  Intersection, Analytic Scape (still), 2025, AI system, camera, video, 10 min, various installations.

• 

  Tree, Analytic Scape (still), 2025, AI system, camera, video, 10 min, various installations.

• 

  Market, Analytic Scape (still), 2025, AI system, camera, video, 10 min, various installations.

• 

  Mountain, Analytic Scape (still), 2025, AI system, camera, video, 10 min, various installations.

• 

  Analytic Scape (installation view at CHANGeUP GROUND, Pohang, 2025), 2025, Single-channel 4K video (8 videos), LED screen (8.5 x 9 m), random loop, sound.

• 

  Analytic Scape (installation view at CHANGeUP GROUND, Pohang, 2025), 2025, Single-channel 4K video (8 videos), LED screen (8.5 x 9 m), random loop, sound.

• 

  Analytic Scape (installation view at CHANGeUP GROUND, Pohang, 2025), 2025, Single-channel 4K video (8 videos), LED screen (8.5 x 9 m), random loop, sound.

* Computer Vision algorithms:
1. YuNet: Face location and size detection.
2. Dlib Face Detection: Face location, size, and landmark detection.
3. DSFD: Face location, size, and confidence detection.
4. Keypoint Communities: Face location, size, landmark, and body pose detection.
5. Facial Expression Recognition with Adaptive Frame Rate based on Multiple Testing Correction: Facial emotion recognition.
6. Distribution-Aware Coordinate Representation for Human Pose Estimation: Human pose estimation.
7. Instances as Queries: Object detection.
8. Densepose: Body pose and body part contour detection.
9. FairFace: Race, gender, and age recognition.
10. L2CS-Net: Gaze direction estimation.
11. Real-time deep hair matting on mobile devices: Hair region segmentation.
12. K-Means Color Quantization: Dominant color detection.
13. Multi-ethnic MEBeauty dataset and facial attractiveness assessment: Facial attractiveness assessment.
14. Distinctive Image Features from Scale-Invariant Keypoints: Feature point detection.
15. Panoptic Feature Pyramid Networks: Object and background segmentation.
16. Mask R-CNN: Instance-level object localization and segmentation.
17. Co‑DETR: Object detection.
18. Lucas-Kanade: Motion detection and tracking.
19. Depth Anything V2: Depth estimation.
20. Open Vocabulary Monocular 3D Object Detection: 3D object detection.