Still dreamin' deeply... .

_梦夢梦夢_梦夢 by @d-pend ^梦夢梦夢^梦夢 . ___ ![afancydd8.jpeg](https://files.peakd.com/file/peakd-hive/d-pend/afYPPxS6-a-fancy-dd8.jpeg)
![asupafance 2.png](https://files.peakd.com/file/peakd-hive/d-pend/ZCyOcD0k-a-supafance202.png) ![asupafance 3.png](https://files.peakd.com/file/peakd-hive/d-pend/M8S4Ts0u-a-supafance203.png)
![asupafance 1.png](https://files.peakd.com/file/peakd-hive/d-pend/HlS7j9yO-a-supafance201.png) ___ ![e.jpeg](https://files.peakd.com/file/peakd-hive/d-pend/ui7IFxQN-e.jpeg) ___ . It's been fun experimenting with Deep Dream Generator recently. As I'm something of an image and video quality aficionado, it's too bad that their highest resolution is still lower than I'd like — but as a futurist it's incredibly exciting for me to consider the future of neural network processing for all kinds of technological applications — which includes digital art. It is one of those uncomfortable (for the traditionalist) and unavoidable truths that the future of human art looks more like a collaboration between the organic and the synthetic than a largely human pursuit. Granted, art has *always* been a collaboration between an artist and the substance of their medium, whether it be sound or stone. But, I digress. . The team of Deep Dream has stated that supporting higher resolutions as well as processing videos is on their roadmap. I really can't wait to be able to incorporate the incredibly bizarre and beautiful imagery into music videos, though to be fair it may actually be a long wait for such an application to be available. I support the project on Patreon so I can process several of the highest resolution images they offer right now per day. I mainly use the "Deep Dream Style" processing in their opt-in newer version which allows for 25 different "layers" of choice when processing. These each have the tendency to produce different structures. To show what I mean I'll put some of DD's example images below: . |Layer 1 example|Layer 2 example|Layer 3 example| |-|-|-| |![layer1example.jpg](https://files.peakd.com/file/peakd-hive/d-pend/5YaS0lLx-layer1example.jpg)|![layer2example.jpg](https://files.peakd.com/file/peakd-hive/d-pend/0W5s81vC-layer2example.jpg)|![layer3example.jpg](https://files.peakd.com/file/peakd-hive/d-pend/S8pOqfoW-layer3example.jpg)| ^{*source — [Deep Dream layer examples page](https://deepdreamgenerator.com/deep-dream-layers-examples)*} . For the images in this post, I decided to process to a "Normal" depth with no "enhance" (which I believe affects the saturation and contrast.) I chose three layers for each image and kept iterating them as follows: *(My source images which where made as photomanipulations from my own photographs are shown first in the table.)* Along with this processing, I made some alterations and artistic adjustments during the creation process using Final Cut Pro. . |![a.jpeg](https://files.peakd.com/file/peakd-hive/d-pend/4bc9gcJE-a.jpeg)|![d 1.jpeg](https://files.peakd.com/file/peakd-hive/d-pend/2MJXEUWj-d201.jpeg)|![e 1.jpeg](https://files.peakd.com/file/peakd-hive/d-pend/X0jCfAoi-e201.jpeg)| |-|-|-| |Layers 11, 13, 19, 11, 13, 19|Layers 2, 20, 22, 2, 20, 22|Layers 17, 14, 4, 17, 14, 4| |![afancydd8.jpg](https://files.peakd.com/file/peakd-hive/d-pend/wC7f53rp-a-fancy-dd8.jpg)|![d6.jpg](https://files.peakd.com/file/peakd-hive/d-pend/xPYcyhoD-d6.jpg)|![e6.jpg](https://files.peakd.com/file/peakd-hive/d-pend/6iyuRVSa-e6.jpg)| . ___ ![e 1.jpeg](https://files.peakd.com/file/peakd-hive/d-pend/vo0zXjyY-e201.jpeg) ___ ![d6.jpeg](https://files.peakd.com/file/peakd-hive/d-pend/XJb2zrVQ-d6.jpeg) ___ `writing and images by @d-pend` created for HIVE on Feb. 26, 2021. _{*images source — personal photographs
processed with Deep Dream and Final Cut Pro.*} ___ ![e6.jpeg](https://files.peakd.com/file/peakd-hive/d-pend/IyviCISq-e6.jpeg)

neuralnetwork

Experts

Show more authors