Embodied Multimodal Multitask Learning

2 minute read

Published:

Summary

  • Deep RL has been successful in training visual navigational agents conditioned on language for multimodal tasks, e.g. semantic goal navigation and embodied question answering
  • Propose a multitask model that can jointly learn these tasks
  • Novel Dual-Attention unit to disentangle words and visual concepts
  • Paper link
  • Success in 3D games such as Doom and DeepmindLab, causing increased interset in training embodied agents, which interact with a 3D environment
    • Receive first-person views
    • Take navigational actions
  • Interseted in specifying goals via language, versus say images or coordinates
    • Allows generalization to new tasks without additional learning
    • Natural medium for humans to communicate with autonomous agents
  • Multimodal tasks require perception from raw pixels, grounding of words to visual attributes, reasoning, navigation, and question answering
  • Multitask approach allows for knowledge to be shared across tasks and increases sample efficiency
  • Most previous work focuses on single tasks or simpler environments, 2D or static.

Background and Methods

  • Consider autonomous agent interacting in episodic environment
    • Receive textual input T at beginning of episode, specifying task
    • State at each time step , is egocentric view
    • Takes action a_t
    • Learn policy to complete task
  • Focus on two visually-grounded language navigation tasks ** Embodied Question Answering: Given a question, gather information from 3D environment to find answer ** Semantic Goal Navigation: Given language instruction, navigate to goal location
  • ViZDoom-based environment
    • Single room with 5 objects, randomized based on textual input
    • 4 actions: forward, left, right, and answer
  • Two difficulty settings
    • Easy: Agent spawned in fixed location with all objects in FOV
    • Hard: Agent and objects spawned randomly
  • BoW and sentence representation (GRU) for textual input, CNN for image
    • Number of CNN channels equals size of vocabulary V
  • Dual-Attention unit combines image and text representations and outputs complete state representation and answer prediction
    • Gated attention between BoW and image features, then sum over channels and softmax to get spatial attention map
    • Spatial attention between image features and spatial attention map
    • Gated attention between spatial attention output and sentence features, then spatial summation and softmax for answer
  • , time step, and task indicator variable fed to policy to produce action
  • Also tried spatial auxiliary task to detect object or attribute in each CNN output channel corresponding to the word

Results

  • +1 reward for reaching correct object or predicting correct answer, -0.001 reward per time step to encourage shorter paths
  • Train for 10 million (Easy) or 50 million frames (Hard)
  • Baselines
    • Image Only
    • Text Only
    • Concat (Hermmann et al. 2017, Misra et al. 2017): semantic goal navigation
    • Gated-Attention (Chaplot et al. 2017): semantic goal navigation
    • FiLM (Perez et al. 2017): question answering
    • PACMAN (Das et al. 2017): question answering
  • Proposed Dual-Attention model outperforms all baselines by a large margin, with and without the auxillary task

Conclusion

  • Model is able to transfer knowledge of words across tasks
  • How to deal with large vocabulary