Christoffer Koo Øhrstrøm
@chrisohrstrom.bsky.social
PhD student at DTU 🇩🇰 Doing research at the intersection of deep learning, event cameras/neuromorphic vision, multi-modal models, and robotics.
We also get a much smaller input sizes with up to a 6.9x reduction over voxels and up to a 8.9x reduction over frames.
November 3, 2025 at 11:44 AM
We also get a much smaller input sizes with up to a 6.9x reduction over voxels and up to a 8.9x reduction over frames.
Results are pretty good. Inference speedups are up to 3.4x over voxels for a point cloud network and up to 10.4x over frames for a Transformer.
This comes without sacrificing accuracy. We even outperform voxels and frames in most cases on gesture recognition and object detection.
This comes without sacrificing accuracy. We even outperform voxels and frames in most cases on gesture recognition and object detection.
November 3, 2025 at 11:44 AM
Results are pretty good. Inference speedups are up to 3.4x over voxels for a point cloud network and up to 10.4x over frames for a Transformer.
This comes without sacrificing accuracy. We even outperform voxels and frames in most cases on gesture recognition and object detection.
This comes without sacrificing accuracy. We even outperform voxels and frames in most cases on gesture recognition and object detection.
Spiking Patches works by creating a grid of patches and let each patch act as spiking neuron. A patch increases its potential whenever an event arrives within the patch, and a token is created everytime a patch spikes (when the potential exceeds a threshold).
November 3, 2025 at 11:44 AM
Spiking Patches works by creating a grid of patches and let each patch act as spiking neuron. A patch increases its potential whenever an event arrives within the patch, and a token is created everytime a patch spikes (when the potential exceeds a threshold).
What if we could represent events (event cameras) in a way that preserves both asynchrony and spatial sparsity?
Exited to share our latest work where we answer this question positively.
Spiking Patches: Asynchronous, Sparse, and Efficient Tokens for Event Cameras
Paper: arxiv.org/abs/2510.26614
Exited to share our latest work where we answer this question positively.
Spiking Patches: Asynchronous, Sparse, and Efficient Tokens for Event Cameras
Paper: arxiv.org/abs/2510.26614
November 3, 2025 at 11:44 AM
What if we could represent events (event cameras) in a way that preserves both asynchrony and spatial sparsity?
Exited to share our latest work where we answer this question positively.
Spiking Patches: Asynchronous, Sparse, and Efficient Tokens for Event Cameras
Paper: arxiv.org/abs/2510.26614
Exited to share our latest work where we answer this question positively.
Spiking Patches: Asynchronous, Sparse, and Efficient Tokens for Event Cameras
Paper: arxiv.org/abs/2510.26614