← Dandy Interview Insights

Dandy·Machine Learning Engineer·Technical Phone Screen·Senior

Senior
May 2026

Summary

ML engineer screen at Dandy with a meaty computer vision problem about height estimation from photos. Single question but they really wanted to see you sweat through the design choices.

Questions Asked (1)

Q1

You have a photo with multiple people in it. The camera's focal length and each person's distance from the camera are unknown. Design an ML system to estimate each person's real-world height.

System DesignTechnical Trade-offsAlgorithms & Data Structures
Author's notes

This one is deceptively open-ended.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the problem constraints and available data, then propose a two-stage approach: first estimate depth or relative scale using monocular cues, then convert to real-world height using a reference object or learned priors. Discuss trade-offs between end-to-end learning and modular pipelines, and emphasize the need for a calibration strategy.

Pro tip: Acknowledge that absolute height estimation from a single image is ill-posed without a reference; propose using known object sizes (e.g., doors, furniture) or human pose priors to resolve scale ambiguity, and suggest a fallback to relative height ranking if absolute values are not critical.

1. Clarify Requirements and Constraints

Ask about the use case: is absolute height needed or relative ranking? What data is available (single image, multiple images, metadata)? What accuracy is acceptable? This shapes the entire design.

2. Choose a Scale Estimation Strategy

Decide between using a reference object of known size, leveraging human pose and anthropometric priors, or training a model to predict depth and focal length jointly. Discuss pros and cons of each.

3. Design the ML Pipeline

Outline a modular pipeline: person detection, pose estimation, depth estimation, and height regression. Alternatively, propose an end-to-end model that takes an image and outputs per-person height, but note data requirements.

4. Address Data and Training

Discuss how to obtain training data: synthetic data with known heights, or real images with ground-truth heights. Mention data augmentation and the challenge of domain shift.

5. Evaluate and Iterate

Define evaluation metrics (MAE, RMSE, rank correlation). Propose validation on a held-out set and discuss potential failure cases (occlusion, unusual poses, extreme distances).

Key Points to Mention

  • Scale ambiguity in monocular images and the need for a reference or prior
  • Use of human pose estimation to get body proportions and anthropometric statistics
  • Depth estimation techniques (monocular depth, stereo, structure from motion) and their limitations
  • Trade-offs between end-to-end learning and modular pipelines (data efficiency, interpretability)
  • Handling multiple people: per-person detection and individual height estimation
  • Evaluation metrics and the importance of a validation set with ground-truth heights

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.