2. Motivation
The subjects/objects are correlated to each other under semantic relationships

3. Contributions
We propose a context-dependent diffusion network (CDDN) framework to deal with visual relationship detection
Semantic graphs to encapsule semantic priors of subjects/objects
Visual scene graphs to capture the surrounding context
Graph diffusion network to learn latent representations of objects

4. Architecture

5. Object Association
Global Semantic Graph： 𝒢 1 = 𝒱 1 , ℰ (Whole training set)
𝓋: object ℯ:
Spatial Scene Graph: 𝒢 2 = 𝒱 2 , ℰ (single picture)
𝓋: bounding box 𝑏 𝑖 , 𝑏 𝑗

6. Diffusion Layer 𝒢= 𝒱,𝐴,𝑋 , 𝒱= 𝑣 1 , 𝑣 2 ,… 𝑣 𝑁 , A∈ ℝ 𝑁×𝑁 ,
𝑋= 𝑥 1 , 𝑥 2 ,… 𝑥 𝑁 ∈ ℝ 𝑁×𝑑
𝐴 ∈ ℝ 𝑁×𝐻×𝑁 : H power series of A
𝑊∈ ℝ 𝑁×𝐻×𝑁

7. Ranking Loss
Triplet 𝑟= 𝑠, 𝑝, 𝑜

8. Experiment