Interview questions in Computer Vision, page 5
Semantic, instance, and panoptic segmentation
Semantic labels every pixel by class without separating objects; instance separates individual objects but may skip background; panoptic unifies both, labeling stuff and distinct thing instances.
Designing a baseline Visual Question Answering model
Encode the image with a CNN, encode the question with an RNN or embedding, fuse the two vectors, and classify over a fixed answer vocabulary.
Camera intrinsics, extrinsics, and the essential matrix
Intrinsics map camera coords to pixels, extrinsics are camera pose in the world; the essential matrix relates normalized points across two views, encoding relative rotation and translation up to scale…
NeRF limitations and advances for robotics
Original NeRF is slow to train and render, per-scene, static, and needs many calibrated views; address speed with explicit grids or Gaussian splatting, dynamics with time-conditioned fields, and scale with…
Formulating a multi-step robot manipulation task
Perception detects and localizes the mug, action space spans navigation and manipulation, and a reward shaped over subgoals (reach, grasp, transport, place) with sparse final success guides learning.
Transfer learning from ResNet50 on small data
Replace the final classification head with one sized to your classes, freeze the pretrained convolutional backbone as a feature extractor, train the new head, then optionally fine-tune top blocks at a low…
Leveraging unlabeled data with limited labels
Pretrain a representation on the million unlabeled images via self-supervision, then fine-tune on the 1,000 labels; or use pseudo-labeling and consistency regularization.
Contrastive self-supervised learning with SimCLR
Two augmentations of one image form a positive pair, other images in the batch are negatives; an encoder plus projection head and the NT-Xent loss pull positives together and push negatives apart.
Prototypical Networks for few-shot classification
An encoder embeds support examples, each class prototype is the mean embedding of its support examples, and a query is classified by nearest prototype using a distance like Euclidean via softmax.
How does MAML's inner and outer loop work?
Inner loop does task-specific gradient steps from shared init, outer loop updates the init for fast adaptability via second-order gradients.
How do BYOL and Barlow Twins avoid representation collapse?
Collapse is embeddings shrinking to a constant or low-rank subspace; BYOL uses predictor plus momentum target plus stop-gradient, Barlow Twins decorrelates feature dimensions.
How do you build an HDR image from bracketed exposures?
Align frames, recover the camera response function, merge to a linear radiance map weighted by exposure, then tone map for display.
Outline the classic image stitching pipeline.
Detect and match features like SIFT, estimate a homography with RANSAC, warp and blend with multiband or feathering.
How does smartphone Portrait Mode produce bokeh?
Estimate per-pixel depth via dual-pixel or stereo or learning, segment the subject, then apply depth-dependent blur.
Compare Gray World and White Patch white balance.
Gray World assumes average scene color is gray, White Patch assumes the brightest pixel is white, both fail on dominant colors or clipping; learning predicts illuminant from data.
How does a plenoptic camera enable post-capture refocus?
Light field records ray direction plus position via a microlens array, refocusing shears and integrates the 4D data to choose a virtual focal plane.
Why merge a burst instead of one long low-light exposure?
Short frames avoid motion blur and clipping while averaging cuts noise; align frames and merge robustly, rejecting misaligned regions to avoid ghosting.
How does focus stacking build an all-in-focus image?
Align frames, measure local sharpness as a focus metric, select or blend the sharpest source per region, then composite.
What data augmentations help small image datasets?
Apply label-preserving transforms like flips, crops, rotation, color jitter, and mixing to enlarge effective data and reduce overfitting.
Precision vs recall in object detection.
Precision is fraction of detections that are correct, recall is fraction of true objects found; prioritize recall for safety-critical detection, precision when false alarms are costly.
We are hiring for this. Every open role lists the topics its interview covers, so you can prepare for the real thing rather than guessing.
See open roles