DeepMind DetCon: Efficient Visual Pretraining with Contrastive Detection | Paper Explained
๐ข SUBSCRIBE TO MY MONTHLY AI NEWSLETTER:
Substack โบ https://aiepiphany.substack.com/
๐จโ๐ฉโ๐งโ๐ฆ JOIN OUR DISCORD COMMUNITY:
Discord โบ https://discord.gg/peBrCpheKE
โค๏ธ Become The AI Epiphany Patreon โค๏ธ โบ https://www.patreon.com/theaiepiphany
In this video, I cover DetCon: Efficient Visual Pretraining with Contrastive Detection - a novel self-supervised method that achieves SOTA results on various transfer learning tasks.
The main idea is to add the semantic segmentation information into the contrastive objective (they used various heuristics to obtain semantic segmentation information inโฆ
Watch on YouTube โ
(saves to browser)
Chapters (9)
News: Discord and AI newsletter!
1:24
Self-supervised learning, BYOL, and SimCLR
3:50
DetCon method overview
11:35
Semantic segmentation heuristics
12:45
Overview of BYOL and SimCLR
20:15
Results
24:10
Impact of segmentation heuristics
27:20
Outro
28:10
Turn on the notification bell, much love!
Playlist
Uploads from Aleksa Gordiฤ - The AI Epiphany ยท Aleksa Gordiฤ - The AI Epiphany ยท 59 of 60
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
โถ
60
Intro | Neural Style Transfer #1
Aleksa Gordiฤ - The AI Epiphany
Basic Theory | Neural Style Transfer #2
Aleksa Gordiฤ - The AI Epiphany
Optimization method | Neural Style Transfer #3
Aleksa Gordiฤ - The AI Epiphany
Advanced Theory | Neural Style Transfer #4
Aleksa Gordiฤ - The AI Epiphany
Anyone can make deepfakes now!
Aleksa Gordiฤ - The AI Epiphany
What is Computer Vision? | The Art of Creating Seeing Machines
Aleksa Gordiฤ - The AI Epiphany
Feed-forward method | Neural Style Transfer #5
Aleksa Gordiฤ - The AI Epiphany
Alan Turing | Computing Machinery and Intelligence
Aleksa Gordiฤ - The AI Epiphany
Feed-forward method (training) | Neural Style Transfer #6
Aleksa Gordiฤ - The AI Epiphany
What is Google Deep Dream? (Basic Theory) | Deep Dream Series #1
Aleksa Gordiฤ - The AI Epiphany
Semantic Segmentation in PyTorch | Neural Style Transfer #7
Aleksa Gordiฤ - The AI Epiphany
How to get started with Machine Learning
Aleksa Gordiฤ - The AI Epiphany
How to learn PyTorch? (3 easy steps) | 2021
Aleksa Gordiฤ - The AI Epiphany
PyTorch or TensorFlow?
Aleksa Gordiฤ - The AI Epiphany
3 Machine Learning Projects For Beginners (Highly visual) | 2021
Aleksa Gordiฤ - The AI Epiphany
Machine Learning Projects (Intermediate level) | 2021
Aleksa Gordiฤ - The AI Epiphany
Cheapest (0$) Deep Learning Hardware Options | 2021
Aleksa Gordiฤ - The AI Epiphany
How to learn deep learning? (Transformers Example)
Aleksa Gordiฤ - The AI Epiphany
How do transformers work? (Attention is all you need)
Aleksa Gordiฤ - The AI Epiphany
Developing a deep learning project (case study on transformer)
Aleksa Gordiฤ - The AI Epiphany
Vision Transformer (ViT) - An image is worth 16x16 words | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
GPT-3 - Language Models are Few-Shot Learners | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Google DeepMind's AlphaFold 2 explained! (Protein folding, AlphaFold 1, a glimpse into AlphaFold 2)
Aleksa Gordiฤ - The AI Epiphany
Attention Is All You Need (Transformer) | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Graph Attention Networks (GAT) | GNN Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Graph Convolutional Networks (GCN) | GNN Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Graph SAGE - Inductive Representation Learning on Large Graphs | GNN Paper Explained
Aleksa Gordiฤ - The AI Epiphany
PinSage - Graph Convolutional Neural Networks for Web-Scale Recommender Systems | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
OpenAI CLIP - Connecting Text and Images | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Temporal Graph Networks (TGN) | GNN Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Graph Neural Network Project Update! (I'm coding GAT from scratch)
Aleksa Gordiฤ - The AI Epiphany
Graph Attention Network Project Walkthrough
Aleksa Gordiฤ - The AI Epiphany
How to get started with Graph ML? (Blog walkthrough)
Aleksa Gordiฤ - The AI Epiphany
DQN - Playing Atari with Deep Reinforcement Learning | RL Paper Explained
Aleksa Gordiฤ - The AI Epiphany
AlphaGo - Mastering the game of Go with deep neural networks and tree search | RL Paper Explained
Aleksa Gordiฤ - The AI Epiphany
DeepMind's AlphaGo Zero and AlphaZero | RL paper explained
Aleksa Gordiฤ - The AI Epiphany
OpenAI - Solving Rubik's Cube with a Robot Hand | RL paper explained
Aleksa Gordiฤ - The AI Epiphany
MuZero - Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model | RL Paper explained
Aleksa Gordiฤ - The AI Epiphany
EfficientNetV2 - Smaller Models and Faster Training | Paper explained
Aleksa Gordiฤ - The AI Epiphany
Implementing DeepMind's DQN from scratch! | Project Update
Aleksa Gordiฤ - The AI Epiphany
MLP-Mixer: An all-MLP Architecture for Vision | Paper explained
Aleksa Gordiฤ - The AI Epiphany
DeepMind's Android RL Environment - AndroidEnv
Aleksa Gordiฤ - The AI Epiphany
When Vision Transformers Outperform ResNets without Pretraining | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Non-Parametric Transformers | Paper explained
Aleksa Gordiฤ - The AI Epiphany
Chip Placement with Deep Reinforcement Learning | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Text Style Brush - Transfer of text aesthetics from a single example | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Graphormer - Do Transformers Really Perform Bad for Graph Representation? | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
GANs N' Roses: Stable, Controllable, Diverse Image to Image Translation | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
VQ-VAEs: Neural Discrete Representation Learning | Paper + PyTorch Code Explained
Aleksa Gordiฤ - The AI Epiphany
VQ-GAN: Taming Transformers for High-Resolution Image Synthesis | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Multimodal Few-Shot Learning with Frozen Language Models | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Focal Transformer: Focal Self-attention for Local-Global Interactions in Vision Transformers
Aleksa Gordiฤ - The AI Epiphany
AudioCLIP: Extending CLIP to Image, Text and Audio | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
RMA: Rapid Motor Adaptation for Legged Robots | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
DALL-E: Zero-Shot Text-to-Image Generation | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Channel Update: vacation, leaving Microsoft, approaching 10k subs and more!
Aleksa Gordiฤ - The AI Epiphany
DETR: End-to-End Object Detection with Transformers | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
DINO: Emerging Properties in Self-Supervised Vision Transformers | Paper Explained!
Aleksa Gordiฤ - The AI Epiphany
DeepMind DetCon: Efficient Visual Pretraining with Contrastive Detection | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
Do Vision Transformers See Like Convolutional Neural Networks? | Paper Explained
Aleksa Gordiฤ - The AI Epiphany
DeepCamp AI