# NAVER AI LAB

<mark style="color:green;">**Introducing NAVER AI LAB**</mark>**&#x20;&#x20;**&#x20;[**Click!**](https://www.youtube.com/watch?v=pDXxvBZ-5tQ)&#x20;

<mark style="color:green;">**Meet Research Scientists at NAVER AI LAB**</mark> <mark style="color:green;"></mark>   [**Click!**](< https://www.youtube.com/watch?v=3-y9mADv8J8>) &#x20;

**\[NAVER AI LAB Job Description]**

* **Research topics**
  * Including, but not limited to -
  * **Next generation of backbones for image, video, and audio recognition.**
    * Multimodal hyperscale representation
    * Novel neural architecture design (e.g., NAS).
    * Object recognition (e.g., classification, detection, segmentation, retrieval, etc.).
    * Lightweight and energy-efficient models (e.g., pruning, quantization, compression).
    * Learning with large-scale insufficient annotations (e.g., weakly- / self- / semi-supervised learning).
    * Novel learning algorithms for NNs (e.g., normalization, optimization, etc.).
  * **Generative models for image, video, text, and audio**&#x20;
    * Uncoditional & conditional image generation
    * Image-to-image and vid-to-vid translation
    * Disentanglement and controllability
    * Cross-modal generation
    * Audio and music generation
    * Effective learning algorithm for generative models
    * Style transfer and super-resolution for images and videos
    * Neural render (NeRF) and super-resolution
  * **Hyperscale language models and their extensions**&#x20;
    * Controllable LM, Hallucination&#x20;
    * Prompt optimization&#x20;
    * Multi-modal and Multi-lingual extension.
    * New evaluation metrics
    * Extension to dialogs, QA, summarization, content generation, etc.
  * **Human computer interaction and interactive AI.**
    * Accessibility
    * Computational Interaction
    * Computational Social Science and Social Computing
    * Data-driven Interface Design
    * Human Computation
    * Visualization
  * **Representation learning for semi-structed or structured data.**
    * Graph representation learning
    * Time-series prediction and representation learning
  * **Trustworthy AI.**
    * Explainable AI and causal inference.
    * Robust machine learning (adversarial robustness, domain generalization).
    * De-biased and fair machine learning.
    * Proper uncertainty estimation (e.g. prediction calibration, probabilistic machine learning).
    * Privacy-preserving AI (e.g. differential privacy, federated learning, etc.).
  * **Audio recognition.**
    * Big representation learning for automatic speech recognition (ASR).
    * Audio-visual speech recognition.
  * **Healthcare AI**
    * EMR/EHR based foundation models (large-scale pre-trained language models) for healthcare
    * Clinical predictive modeling with EMR/EHR (e.g., disease prediction, ICD code mapping)
    * Clinical decision support system
    * Medical image analysis for otorhinolaryngology & dentistry
    * Interpretability of AI models (XAI)
    * Causal inference in machine learning & intervention modeling for healthcare services
  * **Other topics.**
    * AI for social good.
    * Reinforcement learning in the wild.
* **AI Research with External Collaboration**
  * **SNU-NAVER Hyperscale AI Center**&#x20;
    * **Professors: Byung-Gon Chun, Gunhee Kim, Seungwon Hwang, Hyunoh Song, Byoung-Tak Zhang, Taesup Moon, Sang-Goo Lee, Kyomin Jung, Kyoung Mu Lee**
    * **Main topics (not limited)**
      * Advanced hyperscale language models (multimodal, multi-lingual)
      * Reliable and efficient distributed training&#x20;
      * Overcoming limitations of current hyperscale LMs (hallucination, prompt optimization, bias)
      * Advanced large-scale self-supervised learning&#x20;
    * **Some members will contribute as an adjunct professor of SNU.**
  * **KAIST-NAVER Hypercreative AI Center**
    * **Professors: Jaegul Choo, Jinwoo Shin, Sung Ju Hwang, Eunho Yang, Jae-Sik Choi, Juho Lee, Kee-Eung Kim, Alice Oh, Juho Kim, Edward Choi, Minjoon Seo**
    * **Main topics (not limited)**
      * Multi- and cross-modal content generation
      * Generation controllability and quality measurement
      * Representation learning for content generation
    * **Some members will contribute as an adjunct professor of KAIST.**
  * **Academic Advisors**
    * [**Kyunghyun Cho**](https://kyunghyuncho.me/) **(NYU): NLP, hyperscale LM**
    * [**Andrew Zisserman**](https://scholar.google.com/citations?hl=ko\&user=UZ5wscMAAAAJ) **(U. of Oxford): Speech recognition and audio-visual representation learning**
    * [**Jun-Yan Zhu**](https://www.cs.cmu.edu/~junyanz/) **(CMU): Generative models**
    * [**Jonghyun Choi** ](https://ppolon.github.io/)**(GIST): Continual and online learning**
    * [**Joseph Lim**](https://viterbi-web.usc.edu/~limjj/) **(USC): Reinforcement learning in the wild**
* **Requirements**
  * **Research intern**
    * **Experience on research collaborations and paper writing in related fields.**
    * **Proficient programming skills in Python (PyTorch).**&#x20;
    * **Preferred**
      * Currently in an MS or PhD programme in CS, EE, mathematics or other related technical fields.
      * Proficient track record of publications at top-tier conferences in machine learning, computer vision, natural language processing, audio, hci, and speech.
* **Hiring process:**&#x20;
  * **Research intern:**&#x20;
    * Algorithm coding test > Paper implementation or tech talk > Job interview

[Full publication list](https://naver-career.gitbook.io/en/publications/all)&#x20;


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://naver-career.gitbook.io/kr/service/clova/naver-ai-lab.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.