Jean-Marc
@jeanmarcalkazzi.bsky.social
I like to solve meaningful problems. Applied AI @ idealworks , Previously InnovationLab @ BMWGroup.
This looks like a paper worth diving in
"We __prove__ that GNNs, trained to minimize a sparsity-regularized loss over a small set of shortest path instances, exactly implement the Bellman-Ford (BF) algorithm for shortest paths."
"We __prove__ that GNNs, trained to minimize a sparsity-regularized loss over a small set of shortest path instances, exactly implement the Bellman-Ford (BF) algorithm for shortest paths."
April 4, 2025 at 8:16 PM
This looks like a paper worth diving in
"We __prove__ that GNNs, trained to minimize a sparsity-regularized loss over a small set of shortest path instances, exactly implement the Bellman-Ford (BF) algorithm for shortest paths."
"We __prove__ that GNNs, trained to minimize a sparsity-regularized loss over a small set of shortest path instances, exactly implement the Bellman-Ford (BF) algorithm for shortest paths."
What if some agents are slower/faster, how would the reservation table look like?
You discretize time based on their relative speed.
Imagine agent 1 4x as fast as agent 2. The shared table has cells of unit 1. Agent 1 table has cells of unit 2 and Agent 2 table cells of unit 0.5...
You discretize time based on their relative speed.
Imagine agent 1 4x as fast as agent 2. The shared table has cells of unit 1. Agent 1 table has cells of unit 2 and Agent 2 table cells of unit 0.5...
April 4, 2025 at 6:09 AM
What if some agents are slower/faster, how would the reservation table look like?
You discretize time based on their relative speed.
Imagine agent 1 4x as fast as agent 2. The shared table has cells of unit 1. Agent 1 table has cells of unit 2 and Agent 2 table cells of unit 0.5...
You discretize time based on their relative speed.
Imagine agent 1 4x as fast as agent 2. The shared table has cells of unit 1. Agent 1 table has cells of unit 2 and Agent 2 table cells of unit 0.5...
The idea is explained in this figure.
Agents reserve their shortest path in sequence, and a new Space-Time A* approach is used to find the shortest non-reserved path.
Agents reserve their shortest path in sequence, and a new Space-Time A* approach is used to find the shortest non-reserved path.
April 4, 2025 at 6:09 AM
The idea is explained in this figure.
Agents reserve their shortest path in sequence, and a new Space-Time A* approach is used to find the shortest non-reserved path.
Agents reserve their shortest path in sequence, and a new Space-Time A* approach is used to find the shortest non-reserved path.
The idea is very simple, and you "see" the algorithm by looking at this figure
April 3, 2025 at 8:46 PM
The idea is very simple, and you "see" the algorithm by looking at this figure
Why only mention new papers when posting?
Here's CBS, cited *check notes* 1521 times as of today.
"Conflict-based search for optimal multi-agent pathfinding"
Here's CBS, cited *check notes* 1521 times as of today.
"Conflict-based search for optimal multi-agent pathfinding"
April 3, 2025 at 8:46 PM
Why only mention new papers when posting?
Here's CBS, cited *check notes* 1521 times as of today.
"Conflict-based search for optimal multi-agent pathfinding"
Here's CBS, cited *check notes* 1521 times as of today.
"Conflict-based search for optimal multi-agent pathfinding"
And they use POGEMA, beautiful.
March 31, 2025 at 2:53 PM
And they use POGEMA, beautiful.
This figure describes simply the approach taken. Look at the map, predict an action per agent found on the grid, train on SOTA solvers (LaCAM-v1 in that case).
March 31, 2025 at 2:53 PM
This figure describes simply the approach taken. Look at the map, predict an action per agent found on the grid, train on SOTA solvers (LaCAM-v1 in that case).
New paper for the day
"RAILGUN: A Unified Convolutional Policy for Multi-Agent Path Finding Across Different Environments and Tasks"
Interesting part is: "RAILGUN is not an agent-based policy but a map-based policy."
"RAILGUN: A Unified Convolutional Policy for Multi-Agent Path Finding Across Different Environments and Tasks"
Interesting part is: "RAILGUN is not an agent-based policy but a map-based policy."
March 31, 2025 at 2:53 PM
New paper for the day
"RAILGUN: A Unified Convolutional Policy for Multi-Agent Path Finding Across Different Environments and Tasks"
Interesting part is: "RAILGUN is not an agent-based policy but a map-based policy."
"RAILGUN: A Unified Convolutional Policy for Multi-Agent Path Finding Across Different Environments and Tasks"
Interesting part is: "RAILGUN is not an agent-based policy but a map-based policy."
You can find it (open access!) here: ieeexplore.ieee.org/document/105...
It is structured into ML for each of Environment Representation, Solution Planning, and Plan Execution.
It is structured into ML for each of Environment Representation, Solution Planning, and Plan Execution.
March 30, 2025 at 2:51 PM
You can find it (open access!) here: ieeexplore.ieee.org/document/105...
It is structured into ML for each of Environment Representation, Solution Planning, and Plan Execution.
It is structured into ML for each of Environment Representation, Solution Planning, and Plan Execution.
Let's see if posting here about Multi-Agent Systems would get more interactions than Twitter. Starting with a review paper I wrote together with Keisuke Okumura
"A Comprehensive Review on Leveraging Machine Learning for Mutli-Agent Path Finding", published in IEEE Access.
"A Comprehensive Review on Leveraging Machine Learning for Mutli-Agent Path Finding", published in IEEE Access.
March 30, 2025 at 2:51 PM
Let's see if posting here about Multi-Agent Systems would get more interactions than Twitter. Starting with a review paper I wrote together with Keisuke Okumura
"A Comprehensive Review on Leveraging Machine Learning for Mutli-Agent Path Finding", published in IEEE Access.
"A Comprehensive Review on Leveraging Machine Learning for Mutli-Agent Path Finding", published in IEEE Access.
Ants are impressive in swarms.
March 29, 2025 at 1:07 PM
Ants are impressive in swarms.
On the other hand
March 19, 2025 at 7:12 PM
On the other hand
Way better than expected 🤔
March 19, 2025 at 7:06 PM
Way better than expected 🤔